{
  "best_global_step": 5688,
  "best_metric": 0.4171031415462494,
  "best_model_checkpoint": "tiny_bert_km_10_v1_qqp/checkpoint-5688",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 12798,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.6181977987289429,
      "learning_rate": 4.9e-05,
      "loss": 0.494,
      "step": 1422
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.783700222606975,
      "eval_combined_score": 0.7331329164869788,
      "eval_f1": 0.6825656103669825,
      "eval_loss": 0.4529588520526886,
      "eval_runtime": 11.6668,
      "eval_samples_per_second": 3465.375,
      "eval_steps_per_second": 13.543,
      "step": 1422
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.4298322200775146,
      "learning_rate": 4.8e-05,
      "loss": 0.4064,
      "step": 2844
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8001484046500124,
      "eval_combined_score": 0.7520346273891323,
      "eval_f1": 0.7039208501282521,
      "eval_loss": 0.42547887563705444,
      "eval_runtime": 11.3788,
      "eval_samples_per_second": 3553.11,
      "eval_steps_per_second": 13.886,
      "step": 2844
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.6808485984802246,
      "learning_rate": 4.7e-05,
      "loss": 0.3487,
      "step": 4266
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8137768983428147,
      "eval_combined_score": 0.7720519578930665,
      "eval_f1": 0.7303270174433182,
      "eval_loss": 0.4186139702796936,
      "eval_runtime": 11.3626,
      "eval_samples_per_second": 3558.152,
      "eval_steps_per_second": 13.905,
      "step": 4266
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.833254814147949,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.3007,
      "step": 5688
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8211723967350977,
      "eval_combined_score": 0.782321934955684,
      "eval_f1": 0.7434714731762703,
      "eval_loss": 0.4171031415462494,
      "eval_runtime": 11.3784,
      "eval_samples_per_second": 3553.225,
      "eval_steps_per_second": 13.886,
      "step": 5688
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.4554388523101807,
      "learning_rate": 4.5e-05,
      "loss": 0.261,
      "step": 7110
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8216423447934702,
      "eval_combined_score": 0.7901753634624871,
      "eval_f1": 0.758708382131504,
      "eval_loss": 0.4420517385005951,
      "eval_runtime": 11.3653,
      "eval_samples_per_second": 3557.305,
      "eval_steps_per_second": 13.902,
      "step": 7110
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.096228837966919,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.2272,
      "step": 8532
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8227801137768983,
      "eval_combined_score": 0.7922236396978827,
      "eval_f1": 0.761667165618867,
      "eval_loss": 0.45790544152259827,
      "eval_runtime": 11.3923,
      "eval_samples_per_second": 3548.873,
      "eval_steps_per_second": 13.869,
      "step": 8532
    },
    {
      "epoch": 7.0,
      "grad_norm": 4.3414082527160645,
      "learning_rate": 4.3e-05,
      "loss": 0.2,
      "step": 9954
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.827207519168934,
      "eval_combined_score": 0.7969059111279656,
      "eval_f1": 0.7666043030869972,
      "eval_loss": 0.48471689224243164,
      "eval_runtime": 11.3867,
      "eval_samples_per_second": 3550.629,
      "eval_steps_per_second": 13.876,
      "step": 9954
    },
    {
      "epoch": 8.0,
      "grad_norm": 1.463023066520691,
      "learning_rate": 4.2e-05,
      "loss": 0.176,
      "step": 11376
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8280484788523373,
      "eval_combined_score": 0.7961457803919854,
      "eval_f1": 0.7642430819316333,
      "eval_loss": 0.5298035144805908,
      "eval_runtime": 11.3676,
      "eval_samples_per_second": 3556.591,
      "eval_steps_per_second": 13.899,
      "step": 11376
    },
    {
      "epoch": 9.0,
      "grad_norm": 6.6194353103637695,
      "learning_rate": 4.1e-05,
      "loss": 0.1561,
      "step": 12798
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8292604501607717,
      "eval_combined_score": 0.799653043961761,
      "eval_f1": 0.7700456377627503,
      "eval_loss": 0.5476874709129333,
      "eval_runtime": 11.3611,
      "eval_samples_per_second": 3558.634,
      "eval_steps_per_second": 13.907,
      "step": 12798
    },
    {
      "epoch": 9.0,
      "step": 12798,
      "total_flos": 8.587209601564877e+16,
      "train_loss": 0.2855648035406228,
      "train_runtime": 1814.373,
      "train_samples_per_second": 10026.769,
      "train_steps_per_second": 39.187
    }
  ],
  "logging_steps": 1,
  "max_steps": 71100,
  "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": 8.587209601564877e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}