trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.6238698959350586,
  "best_model_checkpoint": "bert_base_km_50_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": 1.111806035041809,
      "learning_rate": 4.9e-05,
      "loss": 0.6192,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6238698959350586,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.6876,
      "eval_samples_per_second": 1516.888,
      "eval_steps_per_second": 7.272,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.4348454475402832,
      "learning_rate": 4.8e-05,
      "loss": 0.5975,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6265467405319214,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.7071,
      "eval_samples_per_second": 1475.092,
      "eval_steps_per_second": 7.071,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.8313347101211548,
      "learning_rate": 4.7e-05,
      "loss": 0.5691,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6778523325920105,
      "eval_loss": 0.6327216625213623,
      "eval_matthews_correlation": 0.02809496484413248,
      "eval_runtime": 0.6947,
      "eval_samples_per_second": 1501.269,
      "eval_steps_per_second": 7.197,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.7352147102355957,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5163,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6883988380432129,
      "eval_loss": 0.6774272918701172,
      "eval_matthews_correlation": 0.12025083258532164,
      "eval_runtime": 0.7181,
      "eval_samples_per_second": 1452.396,
      "eval_steps_per_second": 6.963,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.377678871154785,
      "learning_rate": 4.5e-05,
      "loss": 0.4419,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6558005809783936,
      "eval_loss": 0.7083916068077087,
      "eval_matthews_correlation": 0.09842342220518424,
      "eval_runtime": 0.6793,
      "eval_samples_per_second": 1535.383,
      "eval_steps_per_second": 7.36,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.3255419731140137,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3664,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6116970181465149,
      "eval_loss": 0.8184058666229248,
      "eval_matthews_correlation": 0.047216849138495554,
      "eval_runtime": 0.6952,
      "eval_samples_per_second": 1500.32,
      "eval_steps_per_second": 7.192,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 6749587903150080.0,
      "train_loss": 0.5183892764297187,
      "train_runtime": 99.6293,
      "train_samples_per_second": 4291.406,
      "train_steps_per_second": 17.063
    }
  ],
  "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": 6749587903150080.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}