trainer_state.json

{
  "best_metric": 0.2568,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0992/checkpoints/checkpoint-1665",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1665,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 18.29894256591797,
      "learning_rate": 0.0005,
      "loss": 3.5091,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.15173333333333333,
      "eval_loss": 3.2278530597686768,
      "eval_runtime": 12.0889,
      "eval_samples_per_second": 310.201,
      "eval_steps_per_second": 4.88,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 15.447305679321289,
      "learning_rate": 0.0005,
      "loss": 3.0827,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.1984,
      "eval_loss": 3.017935037612915,
      "eval_runtime": 11.0545,
      "eval_samples_per_second": 339.228,
      "eval_steps_per_second": 5.337,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 14.854424476623535,
      "learning_rate": 0.0005,
      "loss": 2.9293,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.232,
      "eval_loss": 2.884442090988159,
      "eval_runtime": 10.8613,
      "eval_samples_per_second": 345.263,
      "eval_steps_per_second": 5.432,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 17.121829986572266,
      "learning_rate": 0.0005,
      "loss": 2.8076,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.24986666666666665,
      "eval_loss": 2.80948805809021,
      "eval_runtime": 11.1941,
      "eval_samples_per_second": 334.999,
      "eval_steps_per_second": 5.271,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 16.91976547241211,
      "learning_rate": 0.0005,
      "loss": 2.8045,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.2568,
      "eval_loss": 2.7857208251953125,
      "eval_runtime": 13.3717,
      "eval_samples_per_second": 280.442,
      "eval_steps_per_second": 4.412,
      "step": 1665
    }
  ],
  "logging_steps": 500,
  "max_steps": 1665,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 8.2370660311296e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}