trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.6164815425872803,
  "best_model_checkpoint": "bert_base_rand_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": 0.9367762804031372,
      "learning_rate": 4.9e-05,
      "loss": 0.6114,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6164815425872803,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.6761,
      "eval_samples_per_second": 1542.741,
      "eval_steps_per_second": 7.396,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.4945068359375,
      "learning_rate": 4.8e-05,
      "loss": 0.5885,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6883988380432129,
      "eval_loss": 0.6226053833961487,
      "eval_matthews_correlation": 0.11364569650706541,
      "eval_runtime": 0.7607,
      "eval_samples_per_second": 1371.077,
      "eval_steps_per_second": 6.573,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.278257131576538,
      "learning_rate": 4.7e-05,
      "loss": 0.5384,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6701821684837341,
      "eval_loss": 0.6438339352607727,
      "eval_matthews_correlation": 0.09606102313215314,
      "eval_runtime": 0.681,
      "eval_samples_per_second": 1531.562,
      "eval_steps_per_second": 7.342,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.9592695236206055,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.4893,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6644295454025269,
      "eval_loss": 0.7322607636451721,
      "eval_matthews_correlation": 0.09733054506365967,
      "eval_runtime": 0.7117,
      "eval_samples_per_second": 1465.507,
      "eval_steps_per_second": 7.025,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.4407429695129395,
      "learning_rate": 4.5e-05,
      "loss": 0.428,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6567593216896057,
      "eval_loss": 0.712354838848114,
      "eval_matthews_correlation": 0.10027467906851985,
      "eval_runtime": 0.6856,
      "eval_samples_per_second": 1521.368,
      "eval_steps_per_second": 7.293,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.0070488452911377,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3752,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6145733594894409,
      "eval_loss": 0.9127534627914429,
      "eval_matthews_correlation": 0.07810006461499369,
      "eval_runtime": 0.6996,
      "eval_samples_per_second": 1490.881,
      "eval_steps_per_second": 7.147,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 6749587903150080.0,
      "train_loss": 0.5051185757506127,
      "train_runtime": 101.0676,
      "train_samples_per_second": 4230.338,
      "train_steps_per_second": 16.82
    }
  ],
  "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
}