trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.6179393529891968,
  "best_model_checkpoint": "distilbert_km_100_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": 1.0213088989257812,
      "learning_rate": 4.9e-05,
      "loss": 0.6184,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6179393529891968,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.4584,
      "eval_samples_per_second": 2275.26,
      "eval_steps_per_second": 10.907,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.3437625169754028,
      "learning_rate": 4.8e-05,
      "loss": 0.5955,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6893576383590698,
      "eval_loss": 0.629489541053772,
      "eval_matthews_correlation": 0.004051913348362995,
      "eval_runtime": 0.4512,
      "eval_samples_per_second": 2311.392,
      "eval_steps_per_second": 11.081,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.7874633073806763,
      "learning_rate": 4.7e-05,
      "loss": 0.5633,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6816874146461487,
      "eval_loss": 0.6457554697990417,
      "eval_matthews_correlation": 0.04141506615334042,
      "eval_runtime": 0.4366,
      "eval_samples_per_second": 2389.037,
      "eval_steps_per_second": 11.453,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.061436891555786,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5103,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.641418993473053,
      "eval_loss": 0.704453706741333,
      "eval_matthews_correlation": 0.06283033207890582,
      "eval_runtime": 0.4664,
      "eval_samples_per_second": 2236.057,
      "eval_steps_per_second": 10.719,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.2084014415740967,
      "learning_rate": 4.5e-05,
      "loss": 0.4488,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6471716165542603,
      "eval_loss": 0.7053124308586121,
      "eval_matthews_correlation": 0.07524548978135541,
      "eval_runtime": 0.4357,
      "eval_samples_per_second": 2394.029,
      "eval_steps_per_second": 11.477,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.8745715618133545,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3769,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6260786056518555,
      "eval_loss": 0.8186739683151245,
      "eval_matthews_correlation": 0.055260879294738545,
      "eval_runtime": 0.4795,
      "eval_samples_per_second": 2175.278,
      "eval_steps_per_second": 10.428,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 3398186177722368.0,
      "train_loss": 0.5188863978666418,
      "train_runtime": 58.6513,
      "train_samples_per_second": 7289.698,
      "train_steps_per_second": 28.985
    }
  ],
  "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
}