trainer_state.json

{
  "best_metric": 0.0950346514582634,
  "best_model_checkpoint": "vit-large-patch16-224-dungeon-geo-morphs-004/checkpoint-30",
  "epoch": 19.857142857142858,
  "eval_steps": 10,
  "global_step": 30,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 6.571428571428571,
      "grad_norm": 15.927362442016602,
      "learning_rate": 1.555555555555556e-05,
      "loss": 0.7202,
      "step": 10
    },
    {
      "epoch": 6.571428571428571,
      "eval_accuracy": 0.9444444444444444,
      "eval_loss": 0.2301517128944397,
      "eval_runtime": 0.563,
      "eval_samples_per_second": 63.947,
      "eval_steps_per_second": 8.881,
      "step": 10
    },
    {
      "epoch": 13.285714285714286,
      "grad_norm": 6.232780933380127,
      "learning_rate": 8.148148148148148e-06,
      "loss": 0.1296,
      "step": 20
    },
    {
      "epoch": 13.285714285714286,
      "eval_accuracy": 0.9444444444444444,
      "eval_loss": 0.11305957287549973,
      "eval_runtime": 0.6706,
      "eval_samples_per_second": 53.684,
      "eval_steps_per_second": 7.456,
      "step": 20
    },
    {
      "epoch": 19.857142857142858,
      "grad_norm": 4.02134895324707,
      "learning_rate": 7.407407407407407e-07,
      "loss": 0.0385,
      "step": 30
    },
    {
      "epoch": 19.857142857142858,
      "eval_accuracy": 0.9444444444444444,
      "eval_loss": 0.0950346514582634,
      "eval_runtime": 0.5346,
      "eval_samples_per_second": 67.341,
      "eval_steps_per_second": 9.353,
      "step": 30
    },
    {
      "epoch": 19.857142857142858,
      "step": 30,
      "total_flos": 2.6297770138730496e+17,
      "train_loss": 0.2961386869351069,
      "train_runtime": 192.5285,
      "train_samples_per_second": 7.791,
      "train_steps_per_second": 0.156
    }
  ],
  "logging_steps": 10,
  "max_steps": 30,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 30,
  "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": 2.6297770138730496e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}