trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.6189477443695068,
  "best_model_checkpoint": "bert_base_km_50_v1_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": 1.2707092761993408,
      "learning_rate": 4.9e-05,
      "loss": 0.617,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6189477443695068,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.6854,
      "eval_samples_per_second": 1521.716,
      "eval_steps_per_second": 7.295,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.4879816770553589,
      "learning_rate": 4.8e-05,
      "loss": 0.5996,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.693192720413208,
      "eval_loss": 0.6232299208641052,
      "eval_matthews_correlation": 0.0592680243795702,
      "eval_runtime": 0.6773,
      "eval_samples_per_second": 1539.845,
      "eval_steps_per_second": 7.382,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.00919508934021,
      "learning_rate": 4.7e-05,
      "loss": 0.5631,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6605944633483887,
      "eval_loss": 0.6374881267547607,
      "eval_matthews_correlation": -0.0024323816012785963,
      "eval_runtime": 0.6752,
      "eval_samples_per_second": 1544.665,
      "eval_steps_per_second": 7.405,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.3946533203125,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.499,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6711409687995911,
      "eval_loss": 0.6979495882987976,
      "eval_matthews_correlation": 0.05755092158161861,
      "eval_runtime": 0.6813,
      "eval_samples_per_second": 1530.943,
      "eval_steps_per_second": 7.339,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.383981704711914,
      "learning_rate": 4.5e-05,
      "loss": 0.4187,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6558005809783936,
      "eval_loss": 0.7205694913864136,
      "eval_matthews_correlation": 0.11450544212711358,
      "eval_runtime": 0.6667,
      "eval_samples_per_second": 1564.419,
      "eval_steps_per_second": 7.5,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.108778476715088,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3434,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6423777341842651,
      "eval_loss": 0.8469471335411072,
      "eval_matthews_correlation": 0.08517981300248055,
      "eval_runtime": 0.673,
      "eval_samples_per_second": 1549.836,
      "eval_steps_per_second": 7.43,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 6749587903150080.0,
      "train_loss": 0.5067869074204389,
      "train_runtime": 104.3858,
      "train_samples_per_second": 4095.864,
      "train_steps_per_second": 16.286
    }
  ],
  "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
}