{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 25.0,
  "eval_steps": 160,
  "global_step": 800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.32,
      "grad_norm": 0.10831180959939957,
      "learning_rate": 5.625e-07,
      "loss": 0.3009,
      "step": 10
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.143693745136261,
      "learning_rate": 1.1875e-06,
      "loss": 0.4122,
      "step": 20
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.2045338749885559,
      "learning_rate": 1.8125e-06,
      "loss": 0.2875,
      "step": 30
    },
    {
      "epoch": 1.256,
      "grad_norm": 0.13860946893692017,
      "learning_rate": 2.4375e-06,
      "loss": 0.2742,
      "step": 40
    },
    {
      "epoch": 1.576,
      "grad_norm": 0.2552267014980316,
      "learning_rate": 3.0625000000000003e-06,
      "loss": 0.2651,
      "step": 50
    },
    {
      "epoch": 1.896,
      "grad_norm": 0.11148632317781448,
      "learning_rate": 3.6875000000000007e-06,
      "loss": 0.3698,
      "step": 60
    },
    {
      "epoch": 2.192,
      "grad_norm": 0.11400508880615234,
      "learning_rate": 4.312500000000001e-06,
      "loss": 0.2309,
      "step": 70
    },
    {
      "epoch": 2.512,
      "grad_norm": 0.11201531440019608,
      "learning_rate": 4.937500000000001e-06,
      "loss": 0.3079,
      "step": 80
    },
    {
      "epoch": 2.832,
      "grad_norm": 0.1809084713459015,
      "learning_rate": 5.5625000000000005e-06,
      "loss": 0.3236,
      "step": 90
    },
    {
      "epoch": 3.128,
      "grad_norm": 0.074888214468956,
      "learning_rate": 6.1875000000000005e-06,
      "loss": 0.2506,
      "step": 100
    },
    {
      "epoch": 3.448,
      "grad_norm": 0.09450047463178635,
      "learning_rate": 6.8125e-06,
      "loss": 0.2735,
      "step": 110
    },
    {
      "epoch": 3.768,
      "grad_norm": 0.10290012508630753,
      "learning_rate": 7.437500000000001e-06,
      "loss": 0.2252,
      "step": 120
    },
    {
      "epoch": 4.064,
      "grad_norm": 0.14294326305389404,
      "learning_rate": 8.062500000000001e-06,
      "loss": 0.2159,
      "step": 130
    },
    {
      "epoch": 4.384,
      "grad_norm": 0.17784863710403442,
      "learning_rate": 8.687500000000001e-06,
      "loss": 0.2853,
      "step": 140
    },
    {
      "epoch": 4.704,
      "grad_norm": 0.12981906533241272,
      "learning_rate": 9.312500000000001e-06,
      "loss": 0.2709,
      "step": 150
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.11478525400161743,
      "learning_rate": 9.937500000000001e-06,
      "loss": 0.2341,
      "step": 160
    },
    {
      "epoch": 5.32,
      "grad_norm": 0.09994978457689285,
      "learning_rate": 9.999036202410324e-06,
      "loss": 0.2332,
      "step": 170
    },
    {
      "epoch": 5.64,
      "grad_norm": 0.08091111481189728,
      "learning_rate": 9.995705033448435e-06,
      "loss": 0.2412,
      "step": 180
    },
    {
      "epoch": 5.96,
      "grad_norm": 0.17597270011901855,
      "learning_rate": 9.9899961794571e-06,
      "loss": 0.2629,
      "step": 190
    },
    {
      "epoch": 6.256,
      "grad_norm": 0.09343019872903824,
      "learning_rate": 9.981912357541628e-06,
      "loss": 0.2044,
      "step": 200
    },
    {
      "epoch": 6.576,
      "grad_norm": 0.12633207440376282,
      "learning_rate": 9.971457415163435e-06,
      "loss": 0.2328,
      "step": 210
    },
    {
      "epoch": 6.896,
      "grad_norm": 0.2602401077747345,
      "learning_rate": 9.958636328308852e-06,
      "loss": 0.2387,
      "step": 220
    },
    {
      "epoch": 7.192,
      "grad_norm": 0.19835928082466125,
      "learning_rate": 9.943455199120836e-06,
      "loss": 0.2517,
      "step": 230
    },
    {
      "epoch": 7.5120000000000005,
      "grad_norm": 0.09127706289291382,
      "learning_rate": 9.925921252994677e-06,
      "loss": 0.2206,
      "step": 240
    },
    {
      "epoch": 7.832,
      "grad_norm": 0.08452180027961731,
      "learning_rate": 9.90604283513909e-06,
      "loss": 0.2617,
      "step": 250
    },
    {
      "epoch": 8.128,
      "grad_norm": 0.13507255911827087,
      "learning_rate": 9.883829406604363e-06,
      "loss": 0.2002,
      "step": 260
    },
    {
      "epoch": 8.448,
      "grad_norm": 0.09941849857568741,
      "learning_rate": 9.859291539779407e-06,
      "loss": 0.2178,
      "step": 270
    },
    {
      "epoch": 8.768,
      "grad_norm": 0.1347104012966156,
      "learning_rate": 9.83244091335986e-06,
      "loss": 0.1977,
      "step": 280
    },
    {
      "epoch": 9.064,
      "grad_norm": 0.08426441252231598,
      "learning_rate": 9.803290306789676e-06,
      "loss": 0.238,
      "step": 290
    },
    {
      "epoch": 9.384,
      "grad_norm": 0.16955962777137756,
      "learning_rate": 9.771853594178791e-06,
      "loss": 0.226,
      "step": 300
    },
    {
      "epoch": 9.704,
      "grad_norm": 0.14687155187129974,
      "learning_rate": 9.7381457376998e-06,
      "loss": 0.2722,
      "step": 310
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.23694726824760437,
      "learning_rate": 9.702182780466775e-06,
      "loss": 0.1848,
      "step": 320
    },
    {
      "epoch": 10.32,
      "grad_norm": 0.11405057460069656,
      "learning_rate": 9.663981838899612e-06,
      "loss": 0.2289,
      "step": 330
    },
    {
      "epoch": 10.64,
      "grad_norm": 0.06918840110301971,
      "learning_rate": 9.623561094577541e-06,
      "loss": 0.2197,
      "step": 340
    },
    {
      "epoch": 10.96,
      "grad_norm": 0.1575315296649933,
      "learning_rate": 9.58093978558568e-06,
      "loss": 0.1979,
      "step": 350
    },
    {
      "epoch": 11.256,
      "grad_norm": 0.11266372352838516,
      "learning_rate": 9.536138197358747e-06,
      "loss": 0.2036,
      "step": 360
    },
    {
      "epoch": 11.576,
      "grad_norm": 0.15609796345233917,
      "learning_rate": 9.48917765302629e-06,
      "loss": 0.2165,
      "step": 370
    },
    {
      "epoch": 11.896,
      "grad_norm": 0.15501129627227783,
      "learning_rate": 9.440080503264038e-06,
      "loss": 0.225,
      "step": 380
    },
    {
      "epoch": 12.192,
      "grad_norm": 0.14790061116218567,
      "learning_rate": 9.388870115656185e-06,
      "loss": 0.2071,
      "step": 390
    },
    {
      "epoch": 12.512,
      "grad_norm": 0.16440975666046143,
      "learning_rate": 9.335570863573687e-06,
      "loss": 0.2252,
      "step": 400
    },
    {
      "epoch": 12.832,
      "grad_norm": 0.16543489694595337,
      "learning_rate": 9.280208114573859e-06,
      "loss": 0.2614,
      "step": 410
    },
    {
      "epoch": 13.128,
      "grad_norm": 0.15220142900943756,
      "learning_rate": 9.222808218326784e-06,
      "loss": 0.1812,
      "step": 420
    },
    {
      "epoch": 13.448,
      "grad_norm": 0.17462018132209778,
      "learning_rate": 9.163398494074314e-06,
      "loss": 0.237,
      "step": 430
    },
    {
      "epoch": 13.768,
      "grad_norm": 0.21084660291671753,
      "learning_rate": 9.102007217627568e-06,
      "loss": 0.2378,
      "step": 440
    },
    {
      "epoch": 14.064,
      "grad_norm": 0.1916162073612213,
      "learning_rate": 9.038663607909198e-06,
      "loss": 0.1792,
      "step": 450
    },
    {
      "epoch": 14.384,
      "grad_norm": 0.21277697384357452,
      "learning_rate": 8.97339781304675e-06,
      "loss": 0.2217,
      "step": 460
    },
    {
      "epoch": 14.704,
      "grad_norm": 0.12744972109794617,
      "learning_rate": 8.906240896023794e-06,
      "loss": 0.2342,
      "step": 470
    },
    {
      "epoch": 15.0,
      "grad_norm": 0.19419269263744354,
      "learning_rate": 8.837224819895627e-06,
      "loss": 0.2181,
      "step": 480
    },
    {
      "epoch": 15.32,
      "grad_norm": 0.1470198631286621,
      "learning_rate": 8.766382432576589e-06,
      "loss": 0.2157,
      "step": 490
    },
    {
      "epoch": 15.64,
      "grad_norm": 0.1495155543088913,
      "learning_rate": 8.693747451206231e-06,
      "loss": 0.1932,
      "step": 500
    },
    {
      "epoch": 15.96,
      "grad_norm": 0.2851159870624542,
      "learning_rate": 8.61935444610179e-06,
      "loss": 0.207,
      "step": 510
    },
    {
      "epoch": 16.256,
      "grad_norm": 0.13762244582176208,
      "learning_rate": 8.543238824304585e-06,
      "loss": 0.2399,
      "step": 520
    },
    {
      "epoch": 16.576,
      "grad_norm": 0.1461113542318344,
      "learning_rate": 8.465436812728181e-06,
      "loss": 0.216,
      "step": 530
    },
    {
      "epoch": 16.896,
      "grad_norm": 0.10633894801139832,
      "learning_rate": 8.385985440916344e-06,
      "loss": 0.2253,
      "step": 540
    },
    {
      "epoch": 17.192,
      "grad_norm": 0.23198412358760834,
      "learning_rate": 8.304922523418988e-06,
      "loss": 0.1824,
      "step": 550
    },
    {
      "epoch": 17.512,
      "grad_norm": 0.1170201450586319,
      "learning_rate": 8.222286641794488e-06,
      "loss": 0.2284,
      "step": 560
    },
    {
      "epoch": 17.832,
      "grad_norm": 0.12365196645259857,
      "learning_rate": 8.138117126246951e-06,
      "loss": 0.2293,
      "step": 570
    },
    {
      "epoch": 18.128,
      "grad_norm": 0.10224795341491699,
      "learning_rate": 8.052454036907174e-06,
      "loss": 0.1917,
      "step": 580
    },
    {
      "epoch": 18.448,
      "grad_norm": 0.15871191024780273,
      "learning_rate": 7.965338144766186e-06,
      "loss": 0.2351,
      "step": 590
    },
    {
      "epoch": 18.768,
      "grad_norm": 0.16119912266731262,
      "learning_rate": 7.876810912270462e-06,
      "loss": 0.1998,
      "step": 600
    },
    {
      "epoch": 19.064,
      "grad_norm": 0.15410035848617554,
      "learning_rate": 7.786914473588057e-06,
      "loss": 0.2056,
      "step": 610
    },
    {
      "epoch": 19.384,
      "grad_norm": 0.06814954429864883,
      "learning_rate": 7.695691614555002e-06,
      "loss": 0.2218,
      "step": 620
    },
    {
      "epoch": 19.704,
      "grad_norm": 0.11184962838888168,
      "learning_rate": 7.603185752311587e-06,
      "loss": 0.2361,
      "step": 630
    },
    {
      "epoch": 20.0,
      "grad_norm": 1.6558356285095215,
      "learning_rate": 7.50944091463814e-06,
      "loss": 0.2012,
      "step": 640
    },
    {
      "epoch": 20.32,
      "grad_norm": 0.13709917664527893,
      "learning_rate": 7.414501719000187e-06,
      "loss": 0.2123,
      "step": 650
    },
    {
      "epoch": 20.64,
      "grad_norm": 0.1258680671453476,
      "learning_rate": 7.318413351312965e-06,
      "loss": 0.204,
      "step": 660
    },
    {
      "epoch": 20.96,
      "grad_norm": 0.16226519644260406,
      "learning_rate": 7.221221544435364e-06,
      "loss": 0.2101,
      "step": 670
    },
    {
      "epoch": 21.256,
      "grad_norm": 0.12673161923885345,
      "learning_rate": 7.1229725564035665e-06,
      "loss": 0.201,
      "step": 680
    },
    {
      "epoch": 21.576,
      "grad_norm": 0.17159311473369598,
      "learning_rate": 7.023713148414728e-06,
      "loss": 0.1921,
      "step": 690
    },
    {
      "epoch": 21.896,
      "grad_norm": 0.1495276689529419,
      "learning_rate": 6.9234905625711816e-06,
      "loss": 0.2538,
      "step": 700
    },
    {
      "epoch": 22.192,
      "grad_norm": 0.15952152013778687,
      "learning_rate": 6.822352499395751e-06,
      "loss": 0.1912,
      "step": 710
    },
    {
      "epoch": 22.512,
      "grad_norm": 0.1319791078567505,
      "learning_rate": 6.720347095128884e-06,
      "loss": 0.231,
      "step": 720
    },
    {
      "epoch": 22.832,
      "grad_norm": 0.09241311997175217,
      "learning_rate": 6.617522898818426e-06,
      "loss": 0.2196,
      "step": 730
    },
    {
      "epoch": 23.128,
      "grad_norm": 0.1970575898885727,
      "learning_rate": 6.513928849212874e-06,
      "loss": 0.1909,
      "step": 740
    },
    {
      "epoch": 23.448,
      "grad_norm": 0.12065102159976959,
      "learning_rate": 6.4096142514692085e-06,
      "loss": 0.2503,
      "step": 750
    },
    {
      "epoch": 23.768,
      "grad_norm": 0.12692612409591675,
      "learning_rate": 6.304628753686295e-06,
      "loss": 0.1812,
      "step": 760
    },
    {
      "epoch": 24.064,
      "grad_norm": 0.12434836477041245,
      "learning_rate": 6.199022323275083e-06,
      "loss": 0.2124,
      "step": 770
    },
    {
      "epoch": 24.384,
      "grad_norm": 0.17354550957679749,
      "learning_rate": 6.092845223176823e-06,
      "loss": 0.2023,
      "step": 780
    },
    {
      "epoch": 24.704,
      "grad_norm": 0.2949866056442261,
      "learning_rate": 5.986147987940632e-06,
      "loss": 0.2236,
      "step": 790
    },
    {
      "epoch": 25.0,
      "grad_norm": 0.2490314096212387,
      "learning_rate": 5.878981399671774e-06,
      "loss": 0.1977,
      "step": 800
    }
  ],
  "logging_steps": 10,
  "max_steps": 1600,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 160,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.0189850507938365e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}