trainer_state.json

{
  "best_metric": 0.9576,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0703/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.161560297012329,
      "learning_rate": 8.571428571428571e-05,
      "loss": 0.5901,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9245333333333333,
      "eval_loss": 0.2710164785385132,
      "eval_runtime": 11.3929,
      "eval_samples_per_second": 329.153,
      "eval_steps_per_second": 5.179,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.8752124309539795,
      "learning_rate": 7.142857142857143e-05,
      "loss": 0.097,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9384,
      "eval_loss": 0.2201308310031891,
      "eval_runtime": 10.9696,
      "eval_samples_per_second": 341.855,
      "eval_steps_per_second": 5.379,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 9.798035621643066,
      "learning_rate": 5.714285714285714e-05,
      "loss": 0.039,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9461333333333334,
      "eval_loss": 0.22353595495224,
      "eval_runtime": 10.8901,
      "eval_samples_per_second": 344.35,
      "eval_steps_per_second": 5.418,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.03829352557659149,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 0.0118,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9533333333333334,
      "eval_loss": 0.19833391904830933,
      "eval_runtime": 10.4485,
      "eval_samples_per_second": 358.902,
      "eval_steps_per_second": 5.647,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.010190186090767384,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.0051,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9562666666666667,
      "eval_loss": 0.19963541626930237,
      "eval_runtime": 10.3595,
      "eval_samples_per_second": 361.987,
      "eval_steps_per_second": 5.695,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.01567106693983078,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.0024,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9570666666666666,
      "eval_loss": 0.18465691804885864,
      "eval_runtime": 10.496,
      "eval_samples_per_second": 357.28,
      "eval_steps_per_second": 5.621,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 4.2900042900042903e-08,
      "loss": 0.0015,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9576,
      "eval_loss": 0.18887333571910858,
      "eval_runtime": 10.7276,
      "eval_samples_per_second": 349.566,
      "eval_steps_per_second": 5.5,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}