trainer_state.json

{
  "best_metric": 0.659783124923706,
  "best_model_checkpoint": "./results/claim_onlyroberta-base/checkpoint-2699",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 10796,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 13.100720405578613,
      "learning_rate": 4.672437244961049e-05,
      "loss": 0.579,
      "step": 2699
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6555103271521182,
      "eval_conf_mat": [
        [
          4360,
          2480
        ],
        [
          2090,
          4336
        ]
      ],
      "eval_f1": 0.6548859688868751,
      "eval_loss": 0.659783124923706,
      "eval_precision": 0.636150234741784,
      "eval_recall": 0.6747587924058512,
      "eval_runtime": 9.813,
      "eval_samples_per_second": 1351.875,
      "eval_steps_per_second": 42.291,
      "step": 2699
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.234604835510254,
      "learning_rate": 4.338691727463831e-05,
      "loss": 0.4956,
      "step": 5398
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6427709935172622,
      "eval_conf_mat": [
        [
          5881,
          959
        ],
        [
          3780,
          2646
        ]
      ],
      "eval_f1": 0.5275645498953245,
      "eval_loss": 0.8142257928848267,
      "eval_precision": 0.7339805825242719,
      "eval_recall": 0.4117647058823529,
      "eval_runtime": 9.7643,
      "eval_samples_per_second": 1358.627,
      "eval_steps_per_second": 42.502,
      "step": 5398
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.444439888000488,
      "learning_rate": 4.004946209966614e-05,
      "loss": 0.4613,
      "step": 8097
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6129202472486055,
      "eval_conf_mat": [
        [
          5950,
          890
        ],
        [
          4245,
          2181
        ]
      ],
      "eval_f1": 0.45930293776982206,
      "eval_loss": 0.7507337331771851,
      "eval_precision": 0.7101921198306741,
      "eval_recall": 0.33940242763772177,
      "eval_runtime": 9.7395,
      "eval_samples_per_second": 1362.083,
      "eval_steps_per_second": 42.61,
      "step": 8097
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.0880028009414673,
      "learning_rate": 3.6712006924693956e-05,
      "loss": 0.6596,
      "step": 10796
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5156037991858887,
      "eval_conf_mat": [
        [
          6840,
          0
        ],
        [
          6426,
          0
        ]
      ],
      "eval_f1": 0.0,
      "eval_loss": 0.6927554607391357,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 9.7891,
      "eval_samples_per_second": 1355.186,
      "eval_steps_per_second": 42.394,
      "step": 10796
    }
  ],
  "logging_steps": 500,
  "max_steps": 40485,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7272246195180780.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}