{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.125,
  "eval_steps": 500,
  "global_step": 2080,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0,
      "eval_loss": 0.5193641781806946,
      "eval_runtime": 15.6757,
      "eval_samples_per_second": 2.041,
      "eval_steps_per_second": 0.255,
      "step": 0
    },
    {
      "epoch": 0.0006009615384615385,
      "grad_norm": 55.403633140710234,
      "learning_rate": 9e-09,
      "loss": 0.5113,
      "step": 10
    },
    {
      "epoch": 0.001201923076923077,
      "grad_norm": 23.47426110450185,
      "learning_rate": 1.8999999999999998e-08,
      "loss": 0.4385,
      "step": 20
    },
    {
      "epoch": 0.0018028846153846155,
      "grad_norm": 10.224385416254647,
      "learning_rate": 2.9e-08,
      "loss": 0.4586,
      "step": 30
    },
    {
      "epoch": 0.002403846153846154,
      "grad_norm": 20.74032508741799,
      "learning_rate": 3.9e-08,
      "loss": 0.4234,
      "step": 40
    },
    {
      "epoch": 0.0030048076923076925,
      "grad_norm": 21.168707763774822,
      "learning_rate": 4.9e-08,
      "loss": 0.526,
      "step": 50
    },
    {
      "epoch": 0.003605769230769231,
      "grad_norm": 22.648636815908606,
      "learning_rate": 5.899999999999999e-08,
      "loss": 0.5596,
      "step": 60
    },
    {
      "epoch": 0.004206730769230769,
      "grad_norm": 13.508051762249275,
      "learning_rate": 6.900000000000001e-08,
      "loss": 0.4981,
      "step": 70
    },
    {
      "epoch": 0.004807692307692308,
      "grad_norm": 23.403551137207025,
      "learning_rate": 7.899999999999999e-08,
      "loss": 0.5008,
      "step": 80
    },
    {
      "epoch": 0.005408653846153846,
      "grad_norm": 48.6142950712791,
      "learning_rate": 8.899999999999999e-08,
      "loss": 0.4585,
      "step": 90
    },
    {
      "epoch": 0.006009615384615385,
      "grad_norm": 8.884428417822688,
      "learning_rate": 9.9e-08,
      "loss": 0.4388,
      "step": 100
    },
    {
      "epoch": 0.006610576923076923,
      "grad_norm": 9.705307595727518,
      "learning_rate": 1.09e-07,
      "loss": 0.4471,
      "step": 110
    },
    {
      "epoch": 0.007211538461538462,
      "grad_norm": 30.306963311413714,
      "learning_rate": 1.19e-07,
      "loss": 0.4293,
      "step": 120
    },
    {
      "epoch": 0.0078125,
      "grad_norm": 14.67235522750005,
      "learning_rate": 1.29e-07,
      "loss": 0.4955,
      "step": 130
    },
    {
      "epoch": 0.008413461538461538,
      "grad_norm": 41.261340367060896,
      "learning_rate": 1.3900000000000001e-07,
      "loss": 0.5347,
      "step": 140
    },
    {
      "epoch": 0.009014423076923076,
      "grad_norm": 28.145775622131612,
      "learning_rate": 1.49e-07,
      "loss": 0.4671,
      "step": 150
    },
    {
      "epoch": 0.009615384615384616,
      "grad_norm": 13.92342254703078,
      "learning_rate": 1.59e-07,
      "loss": 0.4318,
      "step": 160
    },
    {
      "epoch": 0.010216346153846154,
      "grad_norm": 42.72542828466139,
      "learning_rate": 1.69e-07,
      "loss": 0.4232,
      "step": 170
    },
    {
      "epoch": 0.010817307692307692,
      "grad_norm": 18.910952118646854,
      "learning_rate": 1.7899999999999997e-07,
      "loss": 0.5434,
      "step": 180
    },
    {
      "epoch": 0.01141826923076923,
      "grad_norm": 9.945569021169186,
      "learning_rate": 1.8899999999999999e-07,
      "loss": 0.5091,
      "step": 190
    },
    {
      "epoch": 0.01201923076923077,
      "grad_norm": 12.064577086348633,
      "learning_rate": 1.99e-07,
      "loss": 0.4811,
      "step": 200
    },
    {
      "epoch": 0.012620192307692308,
      "grad_norm": 21.22554429640729,
      "learning_rate": 2.0899999999999998e-07,
      "loss": 0.486,
      "step": 210
    },
    {
      "epoch": 0.013221153846153846,
      "grad_norm": 51.114586227567436,
      "learning_rate": 2.19e-07,
      "loss": 0.5489,
      "step": 220
    },
    {
      "epoch": 0.013822115384615384,
      "grad_norm": 33.767338939816305,
      "learning_rate": 2.29e-07,
      "loss": 0.5006,
      "step": 230
    },
    {
      "epoch": 0.014423076923076924,
      "grad_norm": 20.989645421924674,
      "learning_rate": 2.3899999999999996e-07,
      "loss": 0.5026,
      "step": 240
    },
    {
      "epoch": 0.015024038461538462,
      "grad_norm": 46.94572882286708,
      "learning_rate": 2.4899999999999997e-07,
      "loss": 0.4662,
      "step": 250
    },
    {
      "epoch": 0.015625,
      "grad_norm": 16.1857675192416,
      "learning_rate": 2.59e-07,
      "loss": 0.4428,
      "step": 260
    },
    {
      "epoch": 0.01622596153846154,
      "grad_norm": 22.356923049427976,
      "learning_rate": 2.69e-07,
      "loss": 0.466,
      "step": 270
    },
    {
      "epoch": 0.016826923076923076,
      "grad_norm": 27.919776058404576,
      "learning_rate": 2.79e-07,
      "loss": 0.5299,
      "step": 280
    },
    {
      "epoch": 0.017427884615384616,
      "grad_norm": 20.374817697720214,
      "learning_rate": 2.8899999999999995e-07,
      "loss": 0.5575,
      "step": 290
    },
    {
      "epoch": 0.018028846153846152,
      "grad_norm": 11.1596095155468,
      "learning_rate": 2.9899999999999996e-07,
      "loss": 0.5476,
      "step": 300
    },
    {
      "epoch": 0.018629807692307692,
      "grad_norm": 18.15970681497743,
      "learning_rate": 3.09e-07,
      "loss": 0.521,
      "step": 310
    },
    {
      "epoch": 0.019230769230769232,
      "grad_norm": 7.209018602994909,
      "learning_rate": 3.19e-07,
      "loss": 0.4647,
      "step": 320
    },
    {
      "epoch": 0.019831730769230768,
      "grad_norm": 14.837068682221664,
      "learning_rate": 3.29e-07,
      "loss": 0.4864,
      "step": 330
    },
    {
      "epoch": 0.020432692307692308,
      "grad_norm": 14.235741132745929,
      "learning_rate": 3.39e-07,
      "loss": 0.4107,
      "step": 340
    },
    {
      "epoch": 0.021033653846153848,
      "grad_norm": 27.277651971922182,
      "learning_rate": 3.4899999999999996e-07,
      "loss": 0.4536,
      "step": 350
    },
    {
      "epoch": 0.021634615384615384,
      "grad_norm": 17.10480760395017,
      "learning_rate": 3.5899999999999997e-07,
      "loss": 0.4289,
      "step": 360
    },
    {
      "epoch": 0.022235576923076924,
      "grad_norm": 29.929679141274484,
      "learning_rate": 3.69e-07,
      "loss": 0.4647,
      "step": 370
    },
    {
      "epoch": 0.02283653846153846,
      "grad_norm": 10.398612915945172,
      "learning_rate": 3.79e-07,
      "loss": 0.5167,
      "step": 380
    },
    {
      "epoch": 0.0234375,
      "grad_norm": 9.769808118666157,
      "learning_rate": 3.89e-07,
      "loss": 0.4793,
      "step": 390
    },
    {
      "epoch": 0.02403846153846154,
      "grad_norm": 18.64004530242777,
      "learning_rate": 3.99e-07,
      "loss": 0.4865,
      "step": 400
    },
    {
      "epoch": 0.024639423076923076,
      "grad_norm": 17.699474376401422,
      "learning_rate": 4.0899999999999997e-07,
      "loss": 0.4727,
      "step": 410
    },
    {
      "epoch": 0.025240384615384616,
      "grad_norm": 11.83869387760017,
      "learning_rate": 4.19e-07,
      "loss": 0.4803,
      "step": 420
    },
    {
      "epoch": 0.025841346153846152,
      "grad_norm": 52.6903160822148,
      "learning_rate": 4.29e-07,
      "loss": 0.4587,
      "step": 430
    },
    {
      "epoch": 0.026442307692307692,
      "grad_norm": 8.207347232374364,
      "learning_rate": 4.39e-07,
      "loss": 0.4849,
      "step": 440
    },
    {
      "epoch": 0.027043269230769232,
      "grad_norm": 30.073558497443987,
      "learning_rate": 4.49e-07,
      "loss": 0.5023,
      "step": 450
    },
    {
      "epoch": 0.027644230769230768,
      "grad_norm": 10.373186832780206,
      "learning_rate": 4.59e-07,
      "loss": 0.4538,
      "step": 460
    },
    {
      "epoch": 0.028245192307692308,
      "grad_norm": 25.960320061465094,
      "learning_rate": 4.689999999999999e-07,
      "loss": 0.4712,
      "step": 470
    },
    {
      "epoch": 0.028846153846153848,
      "grad_norm": 12.862144045179527,
      "learning_rate": 4.79e-07,
      "loss": 0.4704,
      "step": 480
    },
    {
      "epoch": 0.029447115384615384,
      "grad_norm": 21.599196101058833,
      "learning_rate": 4.89e-07,
      "loss": 0.4695,
      "step": 490
    },
    {
      "epoch": 0.030048076923076924,
      "grad_norm": 26.447052889694493,
      "learning_rate": 4.99e-07,
      "loss": 0.4574,
      "step": 500
    },
    {
      "epoch": 0.03064903846153846,
      "grad_norm": 19.343061960700307,
      "learning_rate": 5.09e-07,
      "loss": 0.4259,
      "step": 510
    },
    {
      "epoch": 0.03125,
      "grad_norm": 55.39351897669441,
      "learning_rate": 5.19e-07,
      "loss": 0.4362,
      "step": 520
    },
    {
      "epoch": 0.031850961538461536,
      "grad_norm": 9.138204788385957,
      "learning_rate": 5.29e-07,
      "loss": 0.4519,
      "step": 530
    },
    {
      "epoch": 0.03245192307692308,
      "grad_norm": 9.17406599995409,
      "learning_rate": 5.39e-07,
      "loss": 0.4639,
      "step": 540
    },
    {
      "epoch": 0.033052884615384616,
      "grad_norm": 42.607781411989706,
      "learning_rate": 5.490000000000001e-07,
      "loss": 0.4302,
      "step": 550
    },
    {
      "epoch": 0.03365384615384615,
      "grad_norm": 10.368759637182924,
      "learning_rate": 5.590000000000001e-07,
      "loss": 0.4107,
      "step": 560
    },
    {
      "epoch": 0.034254807692307696,
      "grad_norm": 12.31592718573613,
      "learning_rate": 5.69e-07,
      "loss": 0.4403,
      "step": 570
    },
    {
      "epoch": 0.03485576923076923,
      "grad_norm": 13.670395980874881,
      "learning_rate": 5.79e-07,
      "loss": 0.4294,
      "step": 580
    },
    {
      "epoch": 0.03545673076923077,
      "grad_norm": 16.653414422369462,
      "learning_rate": 5.89e-07,
      "loss": 0.439,
      "step": 590
    },
    {
      "epoch": 0.036057692307692304,
      "grad_norm": 10.215056425825546,
      "learning_rate": 5.989999999999999e-07,
      "loss": 0.4069,
      "step": 600
    },
    {
      "epoch": 0.03665865384615385,
      "grad_norm": 31.589724664408692,
      "learning_rate": 6.089999999999999e-07,
      "loss": 0.4354,
      "step": 610
    },
    {
      "epoch": 0.037259615384615384,
      "grad_norm": 21.847510353862578,
      "learning_rate": 6.189999999999999e-07,
      "loss": 0.4638,
      "step": 620
    },
    {
      "epoch": 0.03786057692307692,
      "grad_norm": 8.076334096727454,
      "learning_rate": 6.289999999999999e-07,
      "loss": 0.4312,
      "step": 630
    },
    {
      "epoch": 0.038461538461538464,
      "grad_norm": 22.52385975163401,
      "learning_rate": 6.389999999999999e-07,
      "loss": 0.4364,
      "step": 640
    },
    {
      "epoch": 0.0390625,
      "grad_norm": 13.204088595346095,
      "learning_rate": 6.49e-07,
      "loss": 0.4231,
      "step": 650
    },
    {
      "epoch": 0.039663461538461536,
      "grad_norm": 33.19312446450432,
      "learning_rate": 6.59e-07,
      "loss": 0.4301,
      "step": 660
    },
    {
      "epoch": 0.04026442307692308,
      "grad_norm": 9.715044908453299,
      "learning_rate": 6.69e-07,
      "loss": 0.4201,
      "step": 670
    },
    {
      "epoch": 0.040865384615384616,
      "grad_norm": 10.574912363532471,
      "learning_rate": 6.79e-07,
      "loss": 0.4547,
      "step": 680
    },
    {
      "epoch": 0.04146634615384615,
      "grad_norm": 22.39620841168723,
      "learning_rate": 6.889999999999999e-07,
      "loss": 0.4493,
      "step": 690
    },
    {
      "epoch": 0.042067307692307696,
      "grad_norm": 8.058186609469454,
      "learning_rate": 6.989999999999999e-07,
      "loss": 0.3926,
      "step": 700
    },
    {
      "epoch": 0.04266826923076923,
      "grad_norm": 21.022586695480996,
      "learning_rate": 7.089999999999999e-07,
      "loss": 0.4393,
      "step": 710
    },
    {
      "epoch": 0.04326923076923077,
      "grad_norm": 12.66012747128014,
      "learning_rate": 7.189999999999999e-07,
      "loss": 0.4359,
      "step": 720
    },
    {
      "epoch": 0.043870192307692304,
      "grad_norm": 5.937649569363215,
      "learning_rate": 7.289999999999999e-07,
      "loss": 0.4277,
      "step": 730
    },
    {
      "epoch": 0.04447115384615385,
      "grad_norm": 16.798833074370908,
      "learning_rate": 7.389999999999999e-07,
      "loss": 0.446,
      "step": 740
    },
    {
      "epoch": 0.045072115384615384,
      "grad_norm": 26.4182799886464,
      "learning_rate": 7.489999999999999e-07,
      "loss": 0.4737,
      "step": 750
    },
    {
      "epoch": 0.04567307692307692,
      "grad_norm": 13.91407808362797,
      "learning_rate": 7.59e-07,
      "loss": 0.4796,
      "step": 760
    },
    {
      "epoch": 0.046274038461538464,
      "grad_norm": 16.734955748274775,
      "learning_rate": 7.69e-07,
      "loss": 0.4678,
      "step": 770
    },
    {
      "epoch": 0.046875,
      "grad_norm": 11.89547913614258,
      "learning_rate": 7.79e-07,
      "loss": 0.4536,
      "step": 780
    },
    {
      "epoch": 0.047475961538461536,
      "grad_norm": 16.860857694776378,
      "learning_rate": 7.89e-07,
      "loss": 0.422,
      "step": 790
    },
    {
      "epoch": 0.04807692307692308,
      "grad_norm": 20.294362719824335,
      "learning_rate": 7.99e-07,
      "loss": 0.4675,
      "step": 800
    },
    {
      "epoch": 0.048677884615384616,
      "grad_norm": 32.163806042850815,
      "learning_rate": 8.09e-07,
      "loss": 0.4162,
      "step": 810
    },
    {
      "epoch": 0.04927884615384615,
      "grad_norm": 8.835869696589675,
      "learning_rate": 8.189999999999999e-07,
      "loss": 0.4531,
      "step": 820
    },
    {
      "epoch": 0.049879807692307696,
      "grad_norm": 27.004489539541865,
      "learning_rate": 8.289999999999999e-07,
      "loss": 0.46,
      "step": 830
    },
    {
      "epoch": 0.05048076923076923,
      "grad_norm": 10.039120612814386,
      "learning_rate": 8.389999999999999e-07,
      "loss": 0.4351,
      "step": 840
    },
    {
      "epoch": 0.05108173076923077,
      "grad_norm": 25.890820231883552,
      "learning_rate": 8.489999999999999e-07,
      "loss": 0.4477,
      "step": 850
    },
    {
      "epoch": 0.051682692307692304,
      "grad_norm": 21.84065301973945,
      "learning_rate": 8.59e-07,
      "loss": 0.4028,
      "step": 860
    },
    {
      "epoch": 0.05228365384615385,
      "grad_norm": 17.894672539322194,
      "learning_rate": 8.69e-07,
      "loss": 0.4231,
      "step": 870
    },
    {
      "epoch": 0.052884615384615384,
      "grad_norm": 33.13377776596334,
      "learning_rate": 8.79e-07,
      "loss": 0.4483,
      "step": 880
    },
    {
      "epoch": 0.05348557692307692,
      "grad_norm": 57.77962521420785,
      "learning_rate": 8.89e-07,
      "loss": 0.4528,
      "step": 890
    },
    {
      "epoch": 0.054086538461538464,
      "grad_norm": 14.605643747393765,
      "learning_rate": 8.99e-07,
      "loss": 0.4183,
      "step": 900
    },
    {
      "epoch": 0.0546875,
      "grad_norm": 30.921239622195746,
      "learning_rate": 9.09e-07,
      "loss": 0.4299,
      "step": 910
    },
    {
      "epoch": 0.055288461538461536,
      "grad_norm": 14.136816366550216,
      "learning_rate": 9.19e-07,
      "loss": 0.4105,
      "step": 920
    },
    {
      "epoch": 0.05588942307692308,
      "grad_norm": 11.231190902163389,
      "learning_rate": 9.29e-07,
      "loss": 0.4185,
      "step": 930
    },
    {
      "epoch": 0.056490384615384616,
      "grad_norm": 33.801234798562184,
      "learning_rate": 9.389999999999999e-07,
      "loss": 0.4541,
      "step": 940
    },
    {
      "epoch": 0.05709134615384615,
      "grad_norm": 17.846337242420834,
      "learning_rate": 9.489999999999999e-07,
      "loss": 0.4178,
      "step": 950
    },
    {
      "epoch": 0.057692307692307696,
      "grad_norm": 19.667393686818674,
      "learning_rate": 9.589999999999998e-07,
      "loss": 0.4446,
      "step": 960
    },
    {
      "epoch": 0.05829326923076923,
      "grad_norm": 10.919623781198071,
      "learning_rate": 9.69e-07,
      "loss": 0.4745,
      "step": 970
    },
    {
      "epoch": 0.05889423076923077,
      "grad_norm": 27.87766553645535,
      "learning_rate": 9.789999999999999e-07,
      "loss": 0.4779,
      "step": 980
    },
    {
      "epoch": 0.059495192307692304,
      "grad_norm": 32.8338113927267,
      "learning_rate": 9.89e-07,
      "loss": 0.4962,
      "step": 990
    },
    {
      "epoch": 0.06009615384615385,
      "grad_norm": 54.56857085663595,
      "learning_rate": 9.989999999999999e-07,
      "loss": 0.4523,
      "step": 1000
    },
    {
      "epoch": 0.060697115384615384,
      "grad_norm": 67.84354578257332,
      "learning_rate": 9.994245524296674e-07,
      "loss": 0.3837,
      "step": 1010
    },
    {
      "epoch": 0.06129807692307692,
      "grad_norm": 23.915401977778863,
      "learning_rate": 9.987851662404092e-07,
      "loss": 0.4318,
      "step": 1020
    },
    {
      "epoch": 0.061899038461538464,
      "grad_norm": 18.692398541022605,
      "learning_rate": 9.981457800511507e-07,
      "loss": 0.4429,
      "step": 1030
    },
    {
      "epoch": 0.0625,
      "grad_norm": 8.96899175552455,
      "learning_rate": 9.975063938618924e-07,
      "loss": 0.4265,
      "step": 1040
    },
    {
      "epoch": 0.06310096153846154,
      "grad_norm": 67.76708100008346,
      "learning_rate": 9.968670076726342e-07,
      "loss": 0.4004,
      "step": 1050
    },
    {
      "epoch": 0.06370192307692307,
      "grad_norm": 14.95483642619465,
      "learning_rate": 9.962276214833759e-07,
      "loss": 0.4134,
      "step": 1060
    },
    {
      "epoch": 0.06430288461538461,
      "grad_norm": 11.936205292399153,
      "learning_rate": 9.955882352941176e-07,
      "loss": 0.4206,
      "step": 1070
    },
    {
      "epoch": 0.06490384615384616,
      "grad_norm": 8.144208338195012,
      "learning_rate": 9.949488491048593e-07,
      "loss": 0.4,
      "step": 1080
    },
    {
      "epoch": 0.0655048076923077,
      "grad_norm": 19.659029674044277,
      "learning_rate": 9.94309462915601e-07,
      "loss": 0.3855,
      "step": 1090
    },
    {
      "epoch": 0.06610576923076923,
      "grad_norm": 14.97116772003002,
      "learning_rate": 9.936700767263426e-07,
      "loss": 0.402,
      "step": 1100
    },
    {
      "epoch": 0.06670673076923077,
      "grad_norm": 147.02140982947978,
      "learning_rate": 9.930306905370843e-07,
      "loss": 0.481,
      "step": 1110
    },
    {
      "epoch": 0.0673076923076923,
      "grad_norm": 43.55065984134183,
      "learning_rate": 9.92391304347826e-07,
      "loss": 0.4902,
      "step": 1120
    },
    {
      "epoch": 0.06790865384615384,
      "grad_norm": 7.879535552657374,
      "learning_rate": 9.917519181585678e-07,
      "loss": 0.4542,
      "step": 1130
    },
    {
      "epoch": 0.06850961538461539,
      "grad_norm": 8.151701209609962,
      "learning_rate": 9.911125319693095e-07,
      "loss": 0.414,
      "step": 1140
    },
    {
      "epoch": 0.06911057692307693,
      "grad_norm": 14.637959734877429,
      "learning_rate": 9.904731457800513e-07,
      "loss": 0.3782,
      "step": 1150
    },
    {
      "epoch": 0.06971153846153846,
      "grad_norm": 4.501693875744606,
      "learning_rate": 9.898337595907928e-07,
      "loss": 0.3849,
      "step": 1160
    },
    {
      "epoch": 0.0703125,
      "grad_norm": 35.230763684058516,
      "learning_rate": 9.891943734015345e-07,
      "loss": 0.4404,
      "step": 1170
    },
    {
      "epoch": 0.07091346153846154,
      "grad_norm": 12.442031715529945,
      "learning_rate": 9.885549872122762e-07,
      "loss": 0.4576,
      "step": 1180
    },
    {
      "epoch": 0.07151442307692307,
      "grad_norm": 18.8069021556737,
      "learning_rate": 9.879156010230177e-07,
      "loss": 0.4832,
      "step": 1190
    },
    {
      "epoch": 0.07211538461538461,
      "grad_norm": 15.855907967671428,
      "learning_rate": 9.872762148337595e-07,
      "loss": 0.4352,
      "step": 1200
    },
    {
      "epoch": 0.07271634615384616,
      "grad_norm": 14.369715804290616,
      "learning_rate": 9.866368286445012e-07,
      "loss": 0.4028,
      "step": 1210
    },
    {
      "epoch": 0.0733173076923077,
      "grad_norm": 177.80186681538973,
      "learning_rate": 9.85997442455243e-07,
      "loss": 0.4571,
      "step": 1220
    },
    {
      "epoch": 0.07391826923076923,
      "grad_norm": 11.84969824290971,
      "learning_rate": 9.853580562659845e-07,
      "loss": 0.4467,
      "step": 1230
    },
    {
      "epoch": 0.07451923076923077,
      "grad_norm": 68.42006681557945,
      "learning_rate": 9.847186700767262e-07,
      "loss": 0.4505,
      "step": 1240
    },
    {
      "epoch": 0.0751201923076923,
      "grad_norm": 16.098276335781833,
      "learning_rate": 9.84079283887468e-07,
      "loss": 0.4319,
      "step": 1250
    },
    {
      "epoch": 0.07572115384615384,
      "grad_norm": 16.956837398008382,
      "learning_rate": 9.834398976982096e-07,
      "loss": 0.4748,
      "step": 1260
    },
    {
      "epoch": 0.07632211538461539,
      "grad_norm": 13.890105636051898,
      "learning_rate": 9.828005115089514e-07,
      "loss": 0.4738,
      "step": 1270
    },
    {
      "epoch": 0.07692307692307693,
      "grad_norm": 14.52112888511607,
      "learning_rate": 9.821611253196931e-07,
      "loss": 0.4436,
      "step": 1280
    },
    {
      "epoch": 0.07752403846153846,
      "grad_norm": 14.482587862628632,
      "learning_rate": 9.815217391304348e-07,
      "loss": 0.4213,
      "step": 1290
    },
    {
      "epoch": 0.078125,
      "grad_norm": 15.66954308966829,
      "learning_rate": 9.808823529411764e-07,
      "loss": 0.4889,
      "step": 1300
    },
    {
      "epoch": 0.07872596153846154,
      "grad_norm": 33.3147760442157,
      "learning_rate": 9.80242966751918e-07,
      "loss": 0.4395,
      "step": 1310
    },
    {
      "epoch": 0.07932692307692307,
      "grad_norm": 31.576191017602305,
      "learning_rate": 9.796035805626598e-07,
      "loss": 0.3946,
      "step": 1320
    },
    {
      "epoch": 0.07992788461538461,
      "grad_norm": 15.62093132771886,
      "learning_rate": 9.789641943734016e-07,
      "loss": 0.4729,
      "step": 1330
    },
    {
      "epoch": 0.08052884615384616,
      "grad_norm": 51.855714781154056,
      "learning_rate": 9.783248081841433e-07,
      "loss": 0.4652,
      "step": 1340
    },
    {
      "epoch": 0.0811298076923077,
      "grad_norm": 84.97065453956941,
      "learning_rate": 9.77685421994885e-07,
      "loss": 0.4229,
      "step": 1350
    },
    {
      "epoch": 0.08173076923076923,
      "grad_norm": 25.452407688499903,
      "learning_rate": 9.770460358056265e-07,
      "loss": 0.4064,
      "step": 1360
    },
    {
      "epoch": 0.08233173076923077,
      "grad_norm": 18.263105399640835,
      "learning_rate": 9.764066496163683e-07,
      "loss": 0.4564,
      "step": 1370
    },
    {
      "epoch": 0.0829326923076923,
      "grad_norm": 24.334822911726036,
      "learning_rate": 9.7576726342711e-07,
      "loss": 0.4071,
      "step": 1380
    },
    {
      "epoch": 0.08353365384615384,
      "grad_norm": 11.382524106662059,
      "learning_rate": 9.751278772378515e-07,
      "loss": 0.4332,
      "step": 1390
    },
    {
      "epoch": 0.08413461538461539,
      "grad_norm": 54.06425057509178,
      "learning_rate": 9.744884910485932e-07,
      "loss": 0.4505,
      "step": 1400
    },
    {
      "epoch": 0.08473557692307693,
      "grad_norm": 5.076664593001253,
      "learning_rate": 9.73849104859335e-07,
      "loss": 0.4453,
      "step": 1410
    },
    {
      "epoch": 0.08533653846153846,
      "grad_norm": 21.527761337764254,
      "learning_rate": 9.732097186700767e-07,
      "loss": 0.4719,
      "step": 1420
    },
    {
      "epoch": 0.0859375,
      "grad_norm": 36.50870455739431,
      "learning_rate": 9.725703324808182e-07,
      "loss": 0.4512,
      "step": 1430
    },
    {
      "epoch": 0.08653846153846154,
      "grad_norm": 12.540139357753157,
      "learning_rate": 9.7193094629156e-07,
      "loss": 0.4548,
      "step": 1440
    },
    {
      "epoch": 0.08713942307692307,
      "grad_norm": 15.096306703319206,
      "learning_rate": 9.712915601023017e-07,
      "loss": 0.5067,
      "step": 1450
    },
    {
      "epoch": 0.08774038461538461,
      "grad_norm": 88.89595612518276,
      "learning_rate": 9.706521739130434e-07,
      "loss": 0.469,
      "step": 1460
    },
    {
      "epoch": 0.08834134615384616,
      "grad_norm": 24.05245110004213,
      "learning_rate": 9.700127877237851e-07,
      "loss": 0.4336,
      "step": 1470
    },
    {
      "epoch": 0.0889423076923077,
      "grad_norm": 7.772235228925758,
      "learning_rate": 9.693734015345269e-07,
      "loss": 0.39,
      "step": 1480
    },
    {
      "epoch": 0.08954326923076923,
      "grad_norm": 49.178387681040576,
      "learning_rate": 9.687340153452686e-07,
      "loss": 0.4151,
      "step": 1490
    },
    {
      "epoch": 0.09014423076923077,
      "grad_norm": 9.880025918953056,
      "learning_rate": 9.680946291560101e-07,
      "loss": 0.4437,
      "step": 1500
    },
    {
      "epoch": 0.0907451923076923,
      "grad_norm": 84.04802965246304,
      "learning_rate": 9.674552429667519e-07,
      "loss": 0.4544,
      "step": 1510
    },
    {
      "epoch": 0.09134615384615384,
      "grad_norm": 23.879300812095252,
      "learning_rate": 9.668158567774936e-07,
      "loss": 0.4778,
      "step": 1520
    },
    {
      "epoch": 0.09194711538461539,
      "grad_norm": 8.74377800352843,
      "learning_rate": 9.661764705882353e-07,
      "loss": 0.4424,
      "step": 1530
    },
    {
      "epoch": 0.09254807692307693,
      "grad_norm": 25.340394685074983,
      "learning_rate": 9.65537084398977e-07,
      "loss": 0.4457,
      "step": 1540
    },
    {
      "epoch": 0.09314903846153846,
      "grad_norm": 24.377319461101425,
      "learning_rate": 9.648976982097188e-07,
      "loss": 0.4145,
      "step": 1550
    },
    {
      "epoch": 0.09375,
      "grad_norm": 27.44647420729,
      "learning_rate": 9.642583120204603e-07,
      "loss": 0.3954,
      "step": 1560
    },
    {
      "epoch": 0.09435096153846154,
      "grad_norm": 9.525464831385005,
      "learning_rate": 9.63618925831202e-07,
      "loss": 0.3716,
      "step": 1570
    },
    {
      "epoch": 0.09495192307692307,
      "grad_norm": 27.349989017536107,
      "learning_rate": 9.629795396419438e-07,
      "loss": 0.3668,
      "step": 1580
    },
    {
      "epoch": 0.09555288461538461,
      "grad_norm": 19.197820909162925,
      "learning_rate": 9.623401534526855e-07,
      "loss": 0.4735,
      "step": 1590
    },
    {
      "epoch": 0.09615384615384616,
      "grad_norm": 14.821828935472933,
      "learning_rate": 9.61700767263427e-07,
      "loss": 0.4551,
      "step": 1600
    },
    {
      "epoch": 0.0967548076923077,
      "grad_norm": 10.183698117047507,
      "learning_rate": 9.610613810741687e-07,
      "loss": 0.4463,
      "step": 1610
    },
    {
      "epoch": 0.09735576923076923,
      "grad_norm": 17.003317993340907,
      "learning_rate": 9.604219948849105e-07,
      "loss": 0.4501,
      "step": 1620
    },
    {
      "epoch": 0.09795673076923077,
      "grad_norm": 16.527884738622046,
      "learning_rate": 9.59782608695652e-07,
      "loss": 0.4274,
      "step": 1630
    },
    {
      "epoch": 0.0985576923076923,
      "grad_norm": 96.03639510226907,
      "learning_rate": 9.591432225063937e-07,
      "loss": 0.4002,
      "step": 1640
    },
    {
      "epoch": 0.09915865384615384,
      "grad_norm": 21.560400331772723,
      "learning_rate": 9.585038363171354e-07,
      "loss": 0.3857,
      "step": 1650
    },
    {
      "epoch": 0.09975961538461539,
      "grad_norm": 22.502753235352646,
      "learning_rate": 9.578644501278772e-07,
      "loss": 0.3972,
      "step": 1660
    },
    {
      "epoch": 0.10036057692307693,
      "grad_norm": 13.923835321348774,
      "learning_rate": 9.57225063938619e-07,
      "loss": 0.4161,
      "step": 1670
    },
    {
      "epoch": 0.10096153846153846,
      "grad_norm": 13.89020798215834,
      "learning_rate": 9.565856777493606e-07,
      "loss": 0.4409,
      "step": 1680
    },
    {
      "epoch": 0.1015625,
      "grad_norm": 93.30579652904981,
      "learning_rate": 9.559462915601024e-07,
      "loss": 0.4413,
      "step": 1690
    },
    {
      "epoch": 0.10216346153846154,
      "grad_norm": 10.306967482471663,
      "learning_rate": 9.553069053708439e-07,
      "loss": 0.4306,
      "step": 1700
    },
    {
      "epoch": 0.10276442307692307,
      "grad_norm": 16.35022677130188,
      "learning_rate": 9.546675191815856e-07,
      "loss": 0.4361,
      "step": 1710
    },
    {
      "epoch": 0.10336538461538461,
      "grad_norm": 68.65163238192922,
      "learning_rate": 9.540281329923273e-07,
      "loss": 0.4508,
      "step": 1720
    },
    {
      "epoch": 0.10396634615384616,
      "grad_norm": 14.691319864310291,
      "learning_rate": 9.533887468030691e-07,
      "loss": 0.4637,
      "step": 1730
    },
    {
      "epoch": 0.1045673076923077,
      "grad_norm": 11.655266762752866,
      "learning_rate": 9.527493606138107e-07,
      "loss": 0.4239,
      "step": 1740
    },
    {
      "epoch": 0.10516826923076923,
      "grad_norm": 26.007719217225308,
      "learning_rate": 9.521099744245524e-07,
      "loss": 0.4427,
      "step": 1750
    },
    {
      "epoch": 0.10576923076923077,
      "grad_norm": 31.093138088222684,
      "learning_rate": 9.51470588235294e-07,
      "loss": 0.459,
      "step": 1760
    },
    {
      "epoch": 0.1063701923076923,
      "grad_norm": 8.469416417535411,
      "learning_rate": 9.508312020460358e-07,
      "loss": 0.402,
      "step": 1770
    },
    {
      "epoch": 0.10697115384615384,
      "grad_norm": 16.06555093014836,
      "learning_rate": 9.501918158567774e-07,
      "loss": 0.4276,
      "step": 1780
    },
    {
      "epoch": 0.10757211538461539,
      "grad_norm": 46.21222096469417,
      "learning_rate": 9.495524296675191e-07,
      "loss": 0.4177,
      "step": 1790
    },
    {
      "epoch": 0.10817307692307693,
      "grad_norm": 11.561230412768193,
      "learning_rate": 9.489130434782609e-07,
      "loss": 0.3993,
      "step": 1800
    },
    {
      "epoch": 0.10877403846153846,
      "grad_norm": 33.55933631376849,
      "learning_rate": 9.482736572890026e-07,
      "loss": 0.4206,
      "step": 1810
    },
    {
      "epoch": 0.109375,
      "grad_norm": 11.530474264624262,
      "learning_rate": 9.476342710997442e-07,
      "loss": 0.4378,
      "step": 1820
    },
    {
      "epoch": 0.10997596153846154,
      "grad_norm": 28.092472805389466,
      "learning_rate": 9.469948849104858e-07,
      "loss": 0.4033,
      "step": 1830
    },
    {
      "epoch": 0.11057692307692307,
      "grad_norm": 12.87483056607121,
      "learning_rate": 9.463554987212276e-07,
      "loss": 0.3855,
      "step": 1840
    },
    {
      "epoch": 0.11117788461538461,
      "grad_norm": 12.951129649825642,
      "learning_rate": 9.457161125319693e-07,
      "loss": 0.4156,
      "step": 1850
    },
    {
      "epoch": 0.11177884615384616,
      "grad_norm": 10.238507527434807,
      "learning_rate": 9.450767263427109e-07,
      "loss": 0.4417,
      "step": 1860
    },
    {
      "epoch": 0.1123798076923077,
      "grad_norm": 8.449261275213267,
      "learning_rate": 9.444373401534527e-07,
      "loss": 0.4499,
      "step": 1870
    },
    {
      "epoch": 0.11298076923076923,
      "grad_norm": 33.54997769262099,
      "learning_rate": 9.437979539641944e-07,
      "loss": 0.3718,
      "step": 1880
    },
    {
      "epoch": 0.11358173076923077,
      "grad_norm": 163.06822397421263,
      "learning_rate": 9.43158567774936e-07,
      "loss": 0.3994,
      "step": 1890
    },
    {
      "epoch": 0.1141826923076923,
      "grad_norm": 747.6560259343341,
      "learning_rate": 9.425191815856776e-07,
      "loss": 0.41,
      "step": 1900
    },
    {
      "epoch": 0.11478365384615384,
      "grad_norm": 20.948270060235018,
      "learning_rate": 9.418797953964194e-07,
      "loss": 0.4012,
      "step": 1910
    },
    {
      "epoch": 0.11538461538461539,
      "grad_norm": 16.592630584743493,
      "learning_rate": 9.412404092071611e-07,
      "loss": 0.4071,
      "step": 1920
    },
    {
      "epoch": 0.11598557692307693,
      "grad_norm": 6.4533503474115985,
      "learning_rate": 9.406010230179028e-07,
      "loss": 0.3927,
      "step": 1930
    },
    {
      "epoch": 0.11658653846153846,
      "grad_norm": 9.481676872361602,
      "learning_rate": 9.399616368286445e-07,
      "loss": 0.4486,
      "step": 1940
    },
    {
      "epoch": 0.1171875,
      "grad_norm": 22.879429418595755,
      "learning_rate": 9.393222506393862e-07,
      "loss": 0.4487,
      "step": 1950
    },
    {
      "epoch": 0.11778846153846154,
      "grad_norm": 36.32090030167348,
      "learning_rate": 9.386828644501278e-07,
      "loss": 0.4866,
      "step": 1960
    },
    {
      "epoch": 0.11838942307692307,
      "grad_norm": 15.994462091464001,
      "learning_rate": 9.380434782608695e-07,
      "loss": 0.4853,
      "step": 1970
    },
    {
      "epoch": 0.11899038461538461,
      "grad_norm": 16.476880149018157,
      "learning_rate": 9.374040920716112e-07,
      "loss": 0.4713,
      "step": 1980
    },
    {
      "epoch": 0.11959134615384616,
      "grad_norm": 1316.107870553675,
      "learning_rate": 9.367647058823529e-07,
      "loss": 0.4386,
      "step": 1990
    },
    {
      "epoch": 0.1201923076923077,
      "grad_norm": 20.762433776167253,
      "learning_rate": 9.361253196930946e-07,
      "loss": 0.4609,
      "step": 2000
    },
    {
      "epoch": 0.12079326923076923,
      "grad_norm": 33.600636933197194,
      "learning_rate": 9.354859335038364e-07,
      "loss": 0.4346,
      "step": 2010
    },
    {
      "epoch": 0.12139423076923077,
      "grad_norm": 18.47987496120014,
      "learning_rate": 9.34846547314578e-07,
      "loss": 0.3968,
      "step": 2020
    },
    {
      "epoch": 0.1219951923076923,
      "grad_norm": 12.493741832264165,
      "learning_rate": 9.342071611253196e-07,
      "loss": 0.3862,
      "step": 2030
    },
    {
      "epoch": 0.12259615384615384,
      "grad_norm": 8.016162557709496,
      "learning_rate": 9.335677749360613e-07,
      "loss": 0.3913,
      "step": 2040
    },
    {
      "epoch": 0.12319711538461539,
      "grad_norm": 556.4941548489389,
      "learning_rate": 9.329283887468031e-07,
      "loss": 0.4127,
      "step": 2050
    },
    {
      "epoch": 0.12379807692307693,
      "grad_norm": 25.506919407129853,
      "learning_rate": 9.322890025575447e-07,
      "loss": 0.4349,
      "step": 2060
    },
    {
      "epoch": 0.12439903846153846,
      "grad_norm": 21.228728941085222,
      "learning_rate": 9.316496163682864e-07,
      "loss": 0.4603,
      "step": 2070
    },
    {
      "epoch": 0.125,
      "grad_norm": 9.977415169528852,
      "learning_rate": 9.310102301790282e-07,
      "loss": 0.4158,
      "step": 2080
    },
    {
      "epoch": 0.125,
      "eval_loss": 0.4175701141357422,
      "eval_runtime": 13.8604,
      "eval_samples_per_second": 2.309,
      "eval_steps_per_second": 0.289,
      "step": 2080
    }
  ],
  "logging_steps": 10,
  "max_steps": 16640,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.037184756875264e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}