trainer_state.json

{
  "best_metric": 0.08948543667793274,
  "best_model_checkpoint": "vit-large-patch16-224-dungeon-geo-morphs-004/checkpoint-30",
  "epoch": 20.0,
  "eval_steps": 10,
  "global_step": 35,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 5.714285714285714,
      "grad_norm": 15.572179794311523,
      "learning_rate": 1.6774193548387098e-05,
      "loss": 0.8954,
      "step": 10
    },
    {
      "epoch": 5.714285714285714,
      "eval_accuracy": 0.9444444444444444,
      "eval_loss": 0.3421071469783783,
      "eval_runtime": 0.5133,
      "eval_samples_per_second": 70.129,
      "eval_steps_per_second": 9.74,
      "step": 10
    },
    {
      "epoch": 11.428571428571429,
      "grad_norm": 8.270111083984375,
      "learning_rate": 1.0322580645161291e-05,
      "loss": 0.2087,
      "step": 20
    },
    {
      "epoch": 11.428571428571429,
      "eval_accuracy": 0.9444444444444444,
      "eval_loss": 0.14053164422512054,
      "eval_runtime": 0.5592,
      "eval_samples_per_second": 64.376,
      "eval_steps_per_second": 8.941,
      "step": 20
    },
    {
      "epoch": 17.142857142857142,
      "grad_norm": 6.16154146194458,
      "learning_rate": 3.870967741935484e-06,
      "loss": 0.062,
      "step": 30
    },
    {
      "epoch": 17.142857142857142,
      "eval_accuracy": 0.9444444444444444,
      "eval_loss": 0.08948543667793274,
      "eval_runtime": 0.5801,
      "eval_samples_per_second": 62.053,
      "eval_steps_per_second": 8.619,
      "step": 30
    },
    {
      "epoch": 20.0,
      "step": 35,
      "total_flos": 2.7941380772401152e+17,
      "train_loss": 0.3405552021094731,
      "train_runtime": 211.545,
      "train_samples_per_second": 8.438,
      "train_steps_per_second": 0.165
    }
  ],
  "logging_steps": 10,
  "max_steps": 35,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 35,
  "save_steps": 10,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.01
      },
      "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": 2.7941380772401152e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}