{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0512,
  "eval_steps": 1000,
  "global_step": 800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00064,
      "grad_norm": 1.6144306659698486,
      "learning_rate": 1.1520000000000002e-08,
      "loss": 0.729,
      "step": 10
    },
    {
      "epoch": 0.00128,
      "grad_norm": 2.0952296257019043,
      "learning_rate": 2.4320000000000002e-08,
      "loss": 0.7295,
      "step": 20
    },
    {
      "epoch": 0.00192,
      "grad_norm": 1.3587689399719238,
      "learning_rate": 3.7120000000000004e-08,
      "loss": 0.73,
      "step": 30
    },
    {
      "epoch": 0.00256,
      "grad_norm": 1.2531732320785522,
      "learning_rate": 4.9920000000000006e-08,
      "loss": 0.7221,
      "step": 40
    },
    {
      "epoch": 0.0032,
      "grad_norm": 1.437932014465332,
      "learning_rate": 6.272000000000001e-08,
      "loss": 0.7209,
      "step": 50
    },
    {
      "epoch": 0.00384,
      "grad_norm": 1.418426752090454,
      "learning_rate": 7.552e-08,
      "loss": 0.729,
      "step": 60
    },
    {
      "epoch": 0.00448,
      "grad_norm": 1.9476298093795776,
      "learning_rate": 8.832e-08,
      "loss": 0.7242,
      "step": 70
    },
    {
      "epoch": 0.00512,
      "grad_norm": 1.7948051691055298,
      "learning_rate": 1.0112000000000001e-07,
      "loss": 0.7227,
      "step": 80
    },
    {
      "epoch": 0.00576,
      "grad_norm": 1.6534360647201538,
      "learning_rate": 1.1392e-07,
      "loss": 0.7234,
      "step": 90
    },
    {
      "epoch": 0.0064,
      "grad_norm": 1.0920158624649048,
      "learning_rate": 1.2672e-07,
      "loss": 0.7328,
      "step": 100
    },
    {
      "epoch": 0.00704,
      "grad_norm": 1.977837085723877,
      "learning_rate": 1.3952000000000002e-07,
      "loss": 0.7263,
      "step": 110
    },
    {
      "epoch": 0.00768,
      "grad_norm": 1.388983130455017,
      "learning_rate": 1.5232000000000003e-07,
      "loss": 0.7286,
      "step": 120
    },
    {
      "epoch": 0.00832,
      "grad_norm": 1.2956682443618774,
      "learning_rate": 1.6512e-07,
      "loss": 0.7251,
      "step": 130
    },
    {
      "epoch": 0.00896,
      "grad_norm": 1.8125052452087402,
      "learning_rate": 1.7792e-07,
      "loss": 0.7251,
      "step": 140
    },
    {
      "epoch": 0.0096,
      "grad_norm": 1.626846194267273,
      "learning_rate": 1.9072e-07,
      "loss": 0.727,
      "step": 150
    },
    {
      "epoch": 0.01024,
      "grad_norm": 2.3243086338043213,
      "learning_rate": 2.0352e-07,
      "loss": 0.726,
      "step": 160
    },
    {
      "epoch": 0.01088,
      "grad_norm": 1.4734737873077393,
      "learning_rate": 2.1632e-07,
      "loss": 0.7252,
      "step": 170
    },
    {
      "epoch": 0.01152,
      "grad_norm": 2.090498685836792,
      "learning_rate": 2.2912e-07,
      "loss": 0.7273,
      "step": 180
    },
    {
      "epoch": 0.01216,
      "grad_norm": 1.7563093900680542,
      "learning_rate": 2.4192000000000004e-07,
      "loss": 0.719,
      "step": 190
    },
    {
      "epoch": 0.0128,
      "grad_norm": 1.449843168258667,
      "learning_rate": 2.5472000000000005e-07,
      "loss": 0.7237,
      "step": 200
    },
    {
      "epoch": 0.01344,
      "grad_norm": 144219.625,
      "learning_rate": 5.350742447516642e-07,
      "loss": 0.7218,
      "step": 210
    },
    {
      "epoch": 0.01408,
      "grad_norm": 105046.0234375,
      "learning_rate": 5.606758832565284e-07,
      "loss": 0.718,
      "step": 220
    },
    {
      "epoch": 0.01472,
      "grad_norm": 126142.4296875,
      "learning_rate": 5.862775217613928e-07,
      "loss": 0.7107,
      "step": 230
    },
    {
      "epoch": 0.01536,
      "grad_norm": 92423.2265625,
      "learning_rate": 6.118791602662571e-07,
      "loss": 0.7271,
      "step": 240
    },
    {
      "epoch": 0.016,
      "grad_norm": 98091.828125,
      "learning_rate": 6.374807987711214e-07,
      "loss": 0.7123,
      "step": 250
    },
    {
      "epoch": 0.01664,
      "grad_norm": 131949.578125,
      "learning_rate": 6.630824372759858e-07,
      "loss": 0.7204,
      "step": 260
    },
    {
      "epoch": 0.01728,
      "grad_norm": 112228.5625,
      "learning_rate": 6.8868407578085e-07,
      "loss": 0.722,
      "step": 270
    },
    {
      "epoch": 0.01792,
      "grad_norm": 64587.734375,
      "learning_rate": 7.142857142857143e-07,
      "loss": 0.7263,
      "step": 280
    },
    {
      "epoch": 0.01856,
      "grad_norm": 99893.203125,
      "learning_rate": 7.398873527905787e-07,
      "loss": 0.7169,
      "step": 290
    },
    {
      "epoch": 0.0192,
      "grad_norm": 135749.875,
      "learning_rate": 7.65488991295443e-07,
      "loss": 0.7122,
      "step": 300
    },
    {
      "epoch": 0.01984,
      "grad_norm": 103292.5703125,
      "learning_rate": 7.910906298003073e-07,
      "loss": 0.7183,
      "step": 310
    },
    {
      "epoch": 0.02048,
      "grad_norm": 86927.28125,
      "learning_rate": 8.166922683051716e-07,
      "loss": 0.7192,
      "step": 320
    },
    {
      "epoch": 0.02112,
      "grad_norm": 153738.390625,
      "learning_rate": 8.422939068100359e-07,
      "loss": 0.711,
      "step": 330
    },
    {
      "epoch": 0.02176,
      "grad_norm": 69994.7734375,
      "learning_rate": 8.678955453149002e-07,
      "loss": 0.7176,
      "step": 340
    },
    {
      "epoch": 0.0224,
      "grad_norm": 141370.6875,
      "learning_rate": 8.934971838197646e-07,
      "loss": 0.7105,
      "step": 350
    },
    {
      "epoch": 0.02304,
      "grad_norm": 71139.453125,
      "learning_rate": 9.190988223246289e-07,
      "loss": 0.7126,
      "step": 360
    },
    {
      "epoch": 0.02368,
      "grad_norm": 82039.1953125,
      "learning_rate": 9.447004608294931e-07,
      "loss": 0.7078,
      "step": 370
    },
    {
      "epoch": 0.02432,
      "grad_norm": 71275.7890625,
      "learning_rate": 9.703020993343575e-07,
      "loss": 0.7145,
      "step": 380
    },
    {
      "epoch": 0.02496,
      "grad_norm": 145801.21875,
      "learning_rate": 9.959037378392218e-07,
      "loss": 0.7102,
      "step": 390
    },
    {
      "epoch": 0.0256,
      "grad_norm": 171507.0,
      "learning_rate": 1.021505376344086e-06,
      "loss": 0.7123,
      "step": 400
    },
    {
      "epoch": 0.02624,
      "grad_norm": 79134.203125,
      "learning_rate": 1.0471070148489503e-06,
      "loss": 0.7083,
      "step": 410
    },
    {
      "epoch": 0.02688,
      "grad_norm": 69231.640625,
      "learning_rate": 1.0727086533538148e-06,
      "loss": 0.7105,
      "step": 420
    },
    {
      "epoch": 0.02752,
      "grad_norm": 113099.3984375,
      "learning_rate": 1.0983102918586791e-06,
      "loss": 0.7141,
      "step": 430
    },
    {
      "epoch": 0.02816,
      "grad_norm": 121013.734375,
      "learning_rate": 1.1239119303635434e-06,
      "loss": 0.7146,
      "step": 440
    },
    {
      "epoch": 0.0288,
      "grad_norm": 89184.609375,
      "learning_rate": 1.1495135688684077e-06,
      "loss": 0.7133,
      "step": 450
    },
    {
      "epoch": 0.02944,
      "grad_norm": 176246.890625,
      "learning_rate": 1.175115207373272e-06,
      "loss": 0.7086,
      "step": 460
    },
    {
      "epoch": 0.03008,
      "grad_norm": 88161.2265625,
      "learning_rate": 1.2007168458781362e-06,
      "loss": 0.709,
      "step": 470
    },
    {
      "epoch": 0.03072,
      "grad_norm": 74441.015625,
      "learning_rate": 1.2263184843830007e-06,
      "loss": 0.7023,
      "step": 480
    },
    {
      "epoch": 0.03136,
      "grad_norm": 96409.40625,
      "learning_rate": 1.251920122887865e-06,
      "loss": 0.715,
      "step": 490
    },
    {
      "epoch": 0.032,
      "grad_norm": 81090.6484375,
      "learning_rate": 1.2775217613927293e-06,
      "loss": 0.7109,
      "step": 500
    },
    {
      "epoch": 0.03264,
      "grad_norm": 98153.8828125,
      "learning_rate": 1.3031233998975938e-06,
      "loss": 0.7092,
      "step": 510
    },
    {
      "epoch": 0.03328,
      "grad_norm": 78782.546875,
      "learning_rate": 1.3287250384024578e-06,
      "loss": 0.7048,
      "step": 520
    },
    {
      "epoch": 0.03392,
      "grad_norm": 110360.5,
      "learning_rate": 1.354326676907322e-06,
      "loss": 0.7108,
      "step": 530
    },
    {
      "epoch": 0.03456,
      "grad_norm": 88462.0703125,
      "learning_rate": 1.3799283154121864e-06,
      "loss": 0.7041,
      "step": 540
    },
    {
      "epoch": 0.0352,
      "grad_norm": 97624.7421875,
      "learning_rate": 1.4055299539170509e-06,
      "loss": 0.7114,
      "step": 550
    },
    {
      "epoch": 0.03584,
      "grad_norm": 99471.4375,
      "learning_rate": 1.4311315924219151e-06,
      "loss": 0.7191,
      "step": 560
    },
    {
      "epoch": 0.03648,
      "grad_norm": 79087.90625,
      "learning_rate": 1.4567332309267796e-06,
      "loss": 0.7022,
      "step": 570
    },
    {
      "epoch": 0.03712,
      "grad_norm": 65275.0,
      "learning_rate": 1.4823348694316437e-06,
      "loss": 0.7088,
      "step": 580
    },
    {
      "epoch": 0.03776,
      "grad_norm": 153826.28125,
      "learning_rate": 1.507936507936508e-06,
      "loss": 0.7079,
      "step": 590
    },
    {
      "epoch": 0.0384,
      "grad_norm": 64280.38671875,
      "learning_rate": 1.5335381464413722e-06,
      "loss": 0.7018,
      "step": 600
    },
    {
      "epoch": 0.03904,
      "grad_norm": 65060.80078125,
      "learning_rate": 1.5591397849462367e-06,
      "loss": 0.7027,
      "step": 610
    },
    {
      "epoch": 0.03968,
      "grad_norm": 77339.2890625,
      "learning_rate": 1.584741423451101e-06,
      "loss": 0.7038,
      "step": 620
    },
    {
      "epoch": 0.04032,
      "grad_norm": 123140.5546875,
      "learning_rate": 1.6103430619559655e-06,
      "loss": 0.7019,
      "step": 630
    },
    {
      "epoch": 0.04096,
      "grad_norm": 67502.71875,
      "learning_rate": 1.6359447004608298e-06,
      "loss": 0.7094,
      "step": 640
    },
    {
      "epoch": 0.0416,
      "grad_norm": 95452.1796875,
      "learning_rate": 1.6615463389656938e-06,
      "loss": 0.6998,
      "step": 650
    },
    {
      "epoch": 0.04224,
      "grad_norm": 68556.421875,
      "learning_rate": 1.6871479774705581e-06,
      "loss": 0.694,
      "step": 660
    },
    {
      "epoch": 0.04288,
      "grad_norm": 78265.8046875,
      "learning_rate": 1.7127496159754226e-06,
      "loss": 0.7051,
      "step": 670
    },
    {
      "epoch": 0.04352,
      "grad_norm": 93559.3359375,
      "learning_rate": 1.7383512544802869e-06,
      "loss": 0.6997,
      "step": 680
    },
    {
      "epoch": 0.04416,
      "grad_norm": 88091.9375,
      "learning_rate": 1.7639528929851512e-06,
      "loss": 0.6963,
      "step": 690
    },
    {
      "epoch": 0.0448,
      "grad_norm": 73024.359375,
      "learning_rate": 1.7895545314900157e-06,
      "loss": 0.7021,
      "step": 700
    },
    {
      "epoch": 0.04544,
      "grad_norm": 100058.2890625,
      "learning_rate": 1.8151561699948797e-06,
      "loss": 0.7022,
      "step": 710
    },
    {
      "epoch": 0.04608,
      "grad_norm": 99197.1953125,
      "learning_rate": 1.840757808499744e-06,
      "loss": 0.7017,
      "step": 720
    },
    {
      "epoch": 0.04672,
      "grad_norm": 102018.984375,
      "learning_rate": 1.8663594470046085e-06,
      "loss": 0.6985,
      "step": 730
    },
    {
      "epoch": 0.04736,
      "grad_norm": 101586.0234375,
      "learning_rate": 1.8919610855094728e-06,
      "loss": 0.6991,
      "step": 740
    },
    {
      "epoch": 0.048,
      "grad_norm": 151948.25,
      "learning_rate": 1.9175627240143373e-06,
      "loss": 0.6977,
      "step": 750
    },
    {
      "epoch": 0.04864,
      "grad_norm": 88698.7109375,
      "learning_rate": 1.9431643625192015e-06,
      "loss": 0.6961,
      "step": 760
    },
    {
      "epoch": 0.04928,
      "grad_norm": 82451.9296875,
      "learning_rate": 1.9687660010240654e-06,
      "loss": 0.6898,
      "step": 770
    },
    {
      "epoch": 0.04992,
      "grad_norm": 82236.453125,
      "learning_rate": 1.99436763952893e-06,
      "loss": 0.6886,
      "step": 780
    },
    {
      "epoch": 0.05056,
      "grad_norm": 155064.484375,
      "learning_rate": 2.0199692780337944e-06,
      "loss": 0.6921,
      "step": 790
    },
    {
      "epoch": 0.0512,
      "grad_norm": 72238.6328125,
      "learning_rate": 2.0455709165386586e-06,
      "loss": 0.6932,
      "step": 800
    }
  ],
  "logging_steps": 10,
  "max_steps": 78125,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 200,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 6733455906568320.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}