{
  "best_global_step": 60,
  "best_metric": 0.5067150592803955,
  "best_model_checkpoint": "tinybert_base_train_kd_mrpc/checkpoint-60",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 135,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.0588346719741821,
      "learning_rate": 4.906666666666667e-05,
      "loss": 0.6588,
      "step": 15
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_combined_score": 0.7160919540229884,
      "eval_f1": 0.7655172413793103,
      "eval_loss": 0.6283461451530457,
      "eval_runtime": 0.2187,
      "eval_samples_per_second": 1865.558,
      "eval_steps_per_second": 9.145,
      "step": 15
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.6302603483200073,
      "learning_rate": 4.806666666666667e-05,
      "loss": 0.5998,
      "step": 30
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.696078431372549,
      "eval_combined_score": 0.7508605323006945,
      "eval_f1": 0.8056426332288401,
      "eval_loss": 0.5731440782546997,
      "eval_runtime": 0.2338,
      "eval_samples_per_second": 1744.854,
      "eval_steps_per_second": 8.553,
      "step": 30
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.0050811767578125,
      "learning_rate": 4.706666666666667e-05,
      "loss": 0.5433,
      "step": 45
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7401960784313726,
      "eval_combined_score": 0.786764705882353,
      "eval_f1": 0.8333333333333334,
      "eval_loss": 0.5613774657249451,
      "eval_runtime": 0.2224,
      "eval_samples_per_second": 1834.58,
      "eval_steps_per_second": 8.993,
      "step": 45
    },
    {
      "epoch": 4.0,
      "grad_norm": 10.076091766357422,
      "learning_rate": 4.606666666666667e-05,
      "loss": 0.4574,
      "step": 60
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7916666666666666,
      "eval_combined_score": 0.8239213197969544,
      "eval_f1": 0.856175972927242,
      "eval_loss": 0.5067150592803955,
      "eval_runtime": 0.2267,
      "eval_samples_per_second": 1799.599,
      "eval_steps_per_second": 8.822,
      "step": 60
    },
    {
      "epoch": 5.0,
      "grad_norm": 12.540026664733887,
      "learning_rate": 4.5066666666666667e-05,
      "loss": 0.3525,
      "step": 75
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7573529411764706,
      "eval_combined_score": 0.7888398099711753,
      "eval_f1": 0.8203266787658802,
      "eval_loss": 0.5203803777694702,
      "eval_runtime": 0.2188,
      "eval_samples_per_second": 1864.761,
      "eval_steps_per_second": 9.141,
      "step": 75
    },
    {
      "epoch": 6.0,
      "grad_norm": 10.685439109802246,
      "learning_rate": 4.406666666666667e-05,
      "loss": 0.3015,
      "step": 90
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7426470588235294,
      "eval_combined_score": 0.7904298468231669,
      "eval_f1": 0.8382126348228043,
      "eval_loss": 0.829555332660675,
      "eval_runtime": 0.2251,
      "eval_samples_per_second": 1812.326,
      "eval_steps_per_second": 8.884,
      "step": 90
    },
    {
      "epoch": 7.0,
      "grad_norm": 5.721017360687256,
      "learning_rate": 4.3066666666666665e-05,
      "loss": 0.2978,
      "step": 105
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.75,
      "eval_combined_score": 0.7950626959247649,
      "eval_f1": 0.8401253918495298,
      "eval_loss": 0.7863379120826721,
      "eval_runtime": 0.2298,
      "eval_samples_per_second": 1775.245,
      "eval_steps_per_second": 8.702,
      "step": 105
    },
    {
      "epoch": 8.0,
      "grad_norm": 11.30908203125,
      "learning_rate": 4.206666666666667e-05,
      "loss": 0.2163,
      "step": 120
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.7573529411764706,
      "eval_combined_score": 0.7940610859728507,
      "eval_f1": 0.8307692307692308,
      "eval_loss": 0.7096235752105713,
      "eval_runtime": 0.2225,
      "eval_samples_per_second": 1833.884,
      "eval_steps_per_second": 8.99,
      "step": 120
    },
    {
      "epoch": 9.0,
      "grad_norm": 4.002461910247803,
      "learning_rate": 4.106666666666667e-05,
      "loss": 0.1329,
      "step": 135
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.7549019607843137,
      "eval_combined_score": 0.7881652661064426,
      "eval_f1": 0.8214285714285714,
      "eval_loss": 0.7977427840232849,
      "eval_runtime": 0.2192,
      "eval_samples_per_second": 1861.282,
      "eval_steps_per_second": 9.124,
      "step": 135
    },
    {
      "epoch": 9.0,
      "step": 135,
      "total_flos": 2186506882215936.0,
      "train_loss": 0.3955998191127071,
      "train_runtime": 48.5344,
      "train_samples_per_second": 3778.761,
      "train_steps_per_second": 15.453
    }
  ],
  "logging_steps": 1,
  "max_steps": 750,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 5
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2186506882215936.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}