trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.6146763563156128,
  "best_model_checkpoint": "distilbert_rand_5_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": 1.5647382736206055,
      "learning_rate": 4.9e-05,
      "loss": 0.6165,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6146763563156128,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.4338,
      "eval_samples_per_second": 2404.405,
      "eval_steps_per_second": 11.526,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.6598889827728271,
      "learning_rate": 4.8e-05,
      "loss": 0.5903,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6255093216896057,
      "eval_matthews_correlation": 0.06354272869611655,
      "eval_runtime": 0.4935,
      "eval_samples_per_second": 2113.331,
      "eval_steps_per_second": 10.131,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.00650954246521,
      "learning_rate": 4.7e-05,
      "loss": 0.5486,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6864812970161438,
      "eval_loss": 0.6168355941772461,
      "eval_matthews_correlation": 0.09324875475941957,
      "eval_runtime": 0.5115,
      "eval_samples_per_second": 2039.252,
      "eval_steps_per_second": 9.776,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.9253580570220947,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6692233681678772,
      "eval_loss": 0.694465160369873,
      "eval_matthews_correlation": 0.10919902630524979,
      "eval_runtime": 0.4589,
      "eval_samples_per_second": 2272.973,
      "eval_steps_per_second": 10.896,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.4594075679779053,
      "learning_rate": 4.5e-05,
      "loss": 0.4462,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6768935918807983,
      "eval_loss": 0.6996741890907288,
      "eval_matthews_correlation": 0.1197874632769347,
      "eval_runtime": 0.4437,
      "eval_samples_per_second": 2350.522,
      "eval_steps_per_second": 11.268,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.901954412460327,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3988,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6270374059677124,
      "eval_loss": 0.7782854437828064,
      "eval_matthews_correlation": 0.0704463846695094,
      "eval_runtime": 0.5021,
      "eval_samples_per_second": 2077.322,
      "eval_steps_per_second": 9.958,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 3398186177722368.0,
      "train_loss": 0.5167515885596182,
      "train_runtime": 58.9099,
      "train_samples_per_second": 7257.69,
      "train_steps_per_second": 28.858
    }
  ],
  "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
}