trainer_state.json

{
  "best_metric": 1.0,
  "best_model_checkpoint": "vit-large-patch16-224-in21k-dungeon-geo-morphs-denoised-04Dec24-004/checkpoint-60",
  "epoch": 32.0,
  "eval_steps": 10,
  "global_step": 80,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 4.0,
      "grad_norm": 22.540977478027344,
      "learning_rate": 1.9444444444444445e-05,
      "loss": 1.4428,
      "step": 10
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7191919191919192,
      "eval_loss": 1.0773916244506836,
      "eval_runtime": 6.7476,
      "eval_samples_per_second": 73.36,
      "eval_steps_per_second": 9.188,
      "step": 10
    },
    {
      "epoch": 8.0,
      "grad_norm": 16.13055419921875,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.8313,
      "step": 20
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.901010101010101,
      "eval_loss": 0.5764275789260864,
      "eval_runtime": 7.5966,
      "eval_samples_per_second": 65.161,
      "eval_steps_per_second": 8.162,
      "step": 20
    },
    {
      "epoch": 12.0,
      "grad_norm": 15.618579864501953,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.3864,
      "step": 30
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.9575757575757575,
      "eval_loss": 0.27515658736228943,
      "eval_runtime": 6.9694,
      "eval_samples_per_second": 71.024,
      "eval_steps_per_second": 8.896,
      "step": 30
    },
    {
      "epoch": 16.0,
      "grad_norm": 9.556358337402344,
      "learning_rate": 1.138888888888889e-05,
      "loss": 0.164,
      "step": 40
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.9878787878787879,
      "eval_loss": 0.12990407645702362,
      "eval_runtime": 7.9225,
      "eval_samples_per_second": 62.48,
      "eval_steps_per_second": 7.826,
      "step": 40
    },
    {
      "epoch": 20.0,
      "grad_norm": 3.7325854301452637,
      "learning_rate": 8.611111111111112e-06,
      "loss": 0.08,
      "step": 50
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.9959595959595959,
      "eval_loss": 0.07361849397420883,
      "eval_runtime": 6.9011,
      "eval_samples_per_second": 71.728,
      "eval_steps_per_second": 8.984,
      "step": 50
    },
    {
      "epoch": 24.0,
      "grad_norm": 11.820351600646973,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.046,
      "step": 60
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 1.0,
      "eval_loss": 0.04863843694329262,
      "eval_runtime": 6.5716,
      "eval_samples_per_second": 75.324,
      "eval_steps_per_second": 9.434,
      "step": 60
    },
    {
      "epoch": 28.0,
      "grad_norm": 0.9473109245300293,
      "learning_rate": 3.055555555555556e-06,
      "loss": 0.0243,
      "step": 70
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.9939393939393939,
      "eval_loss": 0.03686562925577164,
      "eval_runtime": 7.118,
      "eval_samples_per_second": 69.542,
      "eval_steps_per_second": 8.71,
      "step": 70
    },
    {
      "epoch": 32.0,
      "grad_norm": 0.8807145357131958,
      "learning_rate": 2.7777777777777776e-07,
      "loss": 0.0194,
      "step": 80
    },
    {
      "epoch": 32.0,
      "eval_accuracy": 1.0,
      "eval_loss": 0.02539004758000374,
      "eval_runtime": 6.5686,
      "eval_samples_per_second": 75.358,
      "eval_steps_per_second": 9.439,
      "step": 80
    },
    {
      "epoch": 32.0,
      "step": 80,
      "total_flos": 7.012786101918106e+17,
      "train_loss": 0.3742759317159653,
      "train_runtime": 515.5392,
      "train_samples_per_second": 6.207,
      "train_steps_per_second": 0.155
    }
  ],
  "logging_steps": 10,
  "max_steps": 80,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 40,
  "save_steps": 10,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.01
      },
      "attributes": {
        "early_stopping_patience_counter": 4
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.012786101918106e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}