{
  "best_global_step": 300,
  "best_metric": 0.997979797979798,
  "best_model_checkpoint": "nci-binary-detector/checkpoint-200",
  "epoch": 0.9803921568627451,
  "eval_steps": 100,
  "global_step": 600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.032679738562091505,
      "grad_norm": 2.849823236465454,
      "learning_rate": 1.2418300653594772e-06,
      "loss": 0.1789,
      "step": 20
    },
    {
      "epoch": 0.06535947712418301,
      "grad_norm": 1.1499906778335571,
      "learning_rate": 2.549019607843137e-06,
      "loss": 0.1045,
      "step": 40
    },
    {
      "epoch": 0.09803921568627451,
      "grad_norm": 1.5816751718521118,
      "learning_rate": 3.856209150326798e-06,
      "loss": 0.0389,
      "step": 60
    },
    {
      "epoch": 0.13071895424836602,
      "grad_norm": 0.3664398789405823,
      "learning_rate": 5.163398692810458e-06,
      "loss": 0.0164,
      "step": 80
    },
    {
      "epoch": 0.16339869281045752,
      "grad_norm": 0.15663352608680725,
      "learning_rate": 6.470588235294119e-06,
      "loss": 0.0123,
      "step": 100
    },
    {
      "epoch": 0.16339869281045752,
      "eval_accuracy": 0.9837869137232195,
      "eval_f1": 0.9860279441117764,
      "eval_loss": 0.005597150418907404,
      "eval_precision": 0.9743589743589743,
      "eval_recall": 0.997979797979798,
      "eval_roc_auc": 0.9985759905705631,
      "eval_runtime": 11.5844,
      "eval_samples_per_second": 149.08,
      "eval_steps_per_second": 4.661,
      "step": 100
    },
    {
      "epoch": 0.19607843137254902,
      "grad_norm": 0.10866066068410873,
      "learning_rate": 7.77777777777778e-06,
      "loss": 0.0073,
      "step": 120
    },
    {
      "epoch": 0.22875816993464052,
      "grad_norm": 0.06542709469795227,
      "learning_rate": 9.084967320261438e-06,
      "loss": 0.0019,
      "step": 140
    },
    {
      "epoch": 0.26143790849673204,
      "grad_norm": 0.15572959184646606,
      "learning_rate": 1.03921568627451e-05,
      "loss": 0.0014,
      "step": 160
    },
    {
      "epoch": 0.29411764705882354,
      "grad_norm": 0.009457970038056374,
      "learning_rate": 1.169934640522876e-05,
      "loss": 0.0066,
      "step": 180
    },
    {
      "epoch": 0.32679738562091504,
      "grad_norm": 0.016104549169540405,
      "learning_rate": 1.3006535947712419e-05,
      "loss": 0.0007,
      "step": 200
    },
    {
      "epoch": 0.32679738562091504,
      "eval_accuracy": 0.9971048060220035,
      "eval_f1": 0.9974785678265254,
      "eval_loss": 0.0025405457708984613,
      "eval_precision": 0.9959718026183283,
      "eval_recall": 0.998989898989899,
      "eval_roc_auc": 0.998836396529748,
      "eval_runtime": 8.7693,
      "eval_samples_per_second": 196.936,
      "eval_steps_per_second": 6.158,
      "step": 200
    },
    {
      "epoch": 0.35947712418300654,
      "grad_norm": 0.10450628399848938,
      "learning_rate": 1.431372549019608e-05,
      "loss": 0.0166,
      "step": 220
    },
    {
      "epoch": 0.39215686274509803,
      "grad_norm": 0.11410112679004669,
      "learning_rate": 1.562091503267974e-05,
      "loss": 0.001,
      "step": 240
    },
    {
      "epoch": 0.42483660130718953,
      "grad_norm": 0.056244976818561554,
      "learning_rate": 1.69281045751634e-05,
      "loss": 0.0044,
      "step": 260
    },
    {
      "epoch": 0.45751633986928103,
      "grad_norm": 0.044668033719062805,
      "learning_rate": 1.823529411764706e-05,
      "loss": 0.0019,
      "step": 280
    },
    {
      "epoch": 0.49019607843137253,
      "grad_norm": 0.004644064232707024,
      "learning_rate": 1.954248366013072e-05,
      "loss": 0.0004,
      "step": 300
    },
    {
      "epoch": 0.49019607843137253,
      "eval_accuracy": 0.9976838448176027,
      "eval_f1": 0.997979797979798,
      "eval_loss": 0.002409732900559902,
      "eval_precision": 0.997979797979798,
      "eval_recall": 0.997979797979798,
      "eval_roc_auc": 0.9996368022148212,
      "eval_runtime": 8.7756,
      "eval_samples_per_second": 196.796,
      "eval_steps_per_second": 6.153,
      "step": 300
    },
    {
      "epoch": 0.5228758169934641,
      "grad_norm": 0.31552189588546753,
      "learning_rate": 1.990559186637618e-05,
      "loss": 0.0033,
      "step": 320
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.0014637637650594115,
      "learning_rate": 1.9760348583877997e-05,
      "loss": 0.0002,
      "step": 340
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 2.498281955718994,
      "learning_rate": 1.9615105301379813e-05,
      "loss": 0.0008,
      "step": 360
    },
    {
      "epoch": 0.6209150326797386,
      "grad_norm": 1.514452014816925e-05,
      "learning_rate": 1.946986201888163e-05,
      "loss": 0.0001,
      "step": 380
    },
    {
      "epoch": 0.6535947712418301,
      "grad_norm": 0.016483725979924202,
      "learning_rate": 1.9324618736383445e-05,
      "loss": 0.0046,
      "step": 400
    },
    {
      "epoch": 0.6535947712418301,
      "eval_accuracy": 0.9936305732484076,
      "eval_f1": 0.994475138121547,
      "eval_loss": 0.0015705447876825929,
      "eval_precision": 0.989010989010989,
      "eval_recall": 1.0,
      "eval_roc_auc": 0.9999013198470458,
      "eval_runtime": 8.774,
      "eval_samples_per_second": 196.831,
      "eval_steps_per_second": 6.155,
      "step": 400
    },
    {
      "epoch": 0.6862745098039216,
      "grad_norm": 0.006017300765961409,
      "learning_rate": 1.9179375453885257e-05,
      "loss": 0.0006,
      "step": 420
    },
    {
      "epoch": 0.7189542483660131,
      "grad_norm": 0.0029072267934679985,
      "learning_rate": 1.9034132171387073e-05,
      "loss": 0.0008,
      "step": 440
    },
    {
      "epoch": 0.7516339869281046,
      "grad_norm": 0.011737746186554432,
      "learning_rate": 1.888888888888889e-05,
      "loss": 0.0001,
      "step": 460
    },
    {
      "epoch": 0.7843137254901961,
      "grad_norm": 0.043070435523986816,
      "learning_rate": 1.8743645606390705e-05,
      "loss": 0.0054,
      "step": 480
    },
    {
      "epoch": 0.8169934640522876,
      "grad_norm": 0.0014130477793514729,
      "learning_rate": 1.859840232389252e-05,
      "loss": 0.0002,
      "step": 500
    },
    {
      "epoch": 0.8169934640522876,
      "eval_accuracy": 0.9976838448176027,
      "eval_f1": 0.997979797979798,
      "eval_loss": 0.0020458686631172895,
      "eval_precision": 0.997979797979798,
      "eval_recall": 0.997979797979798,
      "eval_roc_auc": 0.9999479188081631,
      "eval_runtime": 8.7772,
      "eval_samples_per_second": 196.761,
      "eval_steps_per_second": 6.152,
      "step": 500
    },
    {
      "epoch": 0.8496732026143791,
      "grad_norm": 0.1288619190454483,
      "learning_rate": 1.8453159041394337e-05,
      "loss": 0.0,
      "step": 520
    },
    {
      "epoch": 0.8823529411764706,
      "grad_norm": 0.001577255199663341,
      "learning_rate": 1.8307915758896153e-05,
      "loss": 0.0036,
      "step": 540
    },
    {
      "epoch": 0.9150326797385621,
      "grad_norm": 0.0329805426299572,
      "learning_rate": 1.816267247639797e-05,
      "loss": 0.0034,
      "step": 560
    },
    {
      "epoch": 0.9477124183006536,
      "grad_norm": 0.0058128549717366695,
      "learning_rate": 1.801742919389978e-05,
      "loss": 0.0035,
      "step": 580
    },
    {
      "epoch": 0.9803921568627451,
      "grad_norm": 0.00015560630708932877,
      "learning_rate": 1.7872185911401597e-05,
      "loss": 0.0008,
      "step": 600
    },
    {
      "epoch": 0.9803921568627451,
      "eval_accuracy": 0.9971048060220035,
      "eval_f1": 0.9974785678265254,
      "eval_loss": 0.0012797409435734153,
      "eval_precision": 0.9959718026183283,
      "eval_recall": 0.998989898989899,
      "eval_roc_auc": 0.9999218782122447,
      "eval_runtime": 8.7835,
      "eval_samples_per_second": 196.619,
      "eval_steps_per_second": 6.148,
      "step": 600
    }
  ],
  "logging_steps": 20,
  "max_steps": 3060,
  "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": 3
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6542550643507200.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}