trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.6174131631851196,
  "best_model_checkpoint": "bert_base_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": 1.0261465311050415,
      "learning_rate": 4.9e-05,
      "loss": 0.6144,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6174131631851196,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.7028,
      "eval_samples_per_second": 1484.052,
      "eval_steps_per_second": 7.114,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.3109371662139893,
      "learning_rate": 4.8e-05,
      "loss": 0.5971,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6221110224723816,
      "eval_matthews_correlation": 0.0407377624184135,
      "eval_runtime": 0.6706,
      "eval_samples_per_second": 1555.339,
      "eval_steps_per_second": 7.456,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.4791219234466553,
      "learning_rate": 4.7e-05,
      "loss": 0.566,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6272724270820618,
      "eval_matthews_correlation": 0.10408991954989681,
      "eval_runtime": 0.6992,
      "eval_samples_per_second": 1491.723,
      "eval_steps_per_second": 7.151,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.8909639120101929,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5216,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6586769223213196,
      "eval_loss": 0.6421565413475037,
      "eval_matthews_correlation": 0.059532501362631704,
      "eval_runtime": 0.7216,
      "eval_samples_per_second": 1445.336,
      "eval_steps_per_second": 6.929,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.429867744445801,
      "learning_rate": 4.5e-05,
      "loss": 0.4534,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6673058271408081,
      "eval_loss": 0.69119793176651,
      "eval_matthews_correlation": 0.0983454308058509,
      "eval_runtime": 0.6941,
      "eval_samples_per_second": 1502.723,
      "eval_steps_per_second": 7.204,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.0815577507019043,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3795,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6442952752113342,
      "eval_loss": 0.7572374939918518,
      "eval_matthews_correlation": 0.11568542956406423,
      "eval_runtime": 0.6699,
      "eval_samples_per_second": 1557.023,
      "eval_steps_per_second": 7.464,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 6749587903150080.0,
      "train_loss": 0.5219797807581285,
      "train_runtime": 102.6789,
      "train_samples_per_second": 4163.951,
      "train_steps_per_second": 16.556
    }
  ],
  "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
}