trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.6218016147613525,
  "best_model_checkpoint": "distilbert_km_20_v2_cola/checkpoint-34",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 204,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.0493662357330322,
      "learning_rate": 4.9e-05,
      "loss": 0.617,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6218016147613525,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.4438,
      "eval_samples_per_second": 2349.979,
      "eval_steps_per_second": 11.265,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.4519679546356201,
      "learning_rate": 4.8e-05,
      "loss": 0.5982,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6874400973320007,
      "eval_loss": 0.62346351146698,
      "eval_matthews_correlation": -0.02339059069282274,
      "eval_runtime": 0.4557,
      "eval_samples_per_second": 2288.988,
      "eval_steps_per_second": 10.973,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.4409375190734863,
      "learning_rate": 4.7e-05,
      "loss": 0.5656,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6740172505378723,
      "eval_loss": 0.6231377124786377,
      "eval_matthews_correlation": 0.012447227759841093,
      "eval_runtime": 0.4899,
      "eval_samples_per_second": 2129.049,
      "eval_steps_per_second": 10.206,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.2414803504943848,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5134,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6519654989242554,
      "eval_loss": 0.6778802275657654,
      "eval_matthews_correlation": 0.05913990204236021,
      "eval_runtime": 0.4375,
      "eval_samples_per_second": 2383.897,
      "eval_steps_per_second": 11.428,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.8618477582931519,
      "learning_rate": 4.5e-05,
      "loss": 0.4471,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6605944633483887,
      "eval_loss": 0.6931583285331726,
      "eval_matthews_correlation": 0.09451155237608601,
      "eval_runtime": 0.4739,
      "eval_samples_per_second": 2200.883,
      "eval_steps_per_second": 10.551,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.7331478595733643,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3863,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6538830399513245,
      "eval_loss": 0.7930698990821838,
      "eval_matthews_correlation": 0.08801919690309869,
      "eval_runtime": 0.4918,
      "eval_samples_per_second": 2120.801,
      "eval_steps_per_second": 10.167,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 3398186177722368.0,
      "train_loss": 0.5212595743291518,
      "train_runtime": 58.5621,
      "train_samples_per_second": 7300.799,
      "train_steps_per_second": 29.029
    }
  ],
  "logging_steps": 1,
  "max_steps": 1700,
  "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": 3398186177722368.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}