trainer_state.json

{
  "best_metric": 0.8650666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0913/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": Infinity,
      "learning_rate": 2.7151869176294504e-05,
      "loss": 2.2705,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7464,
      "eval_loss": 1.138380765914917,
      "eval_runtime": 6.3197,
      "eval_samples_per_second": 593.379,
      "eval_steps_per_second": 9.336,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 1.968905659703686e-05,
      "loss": 0.7785,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8141333333333334,
      "eval_loss": 0.6787221431732178,
      "eval_runtime": 6.2613,
      "eval_samples_per_second": 598.914,
      "eval_steps_per_second": 9.423,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 38.84733963012695,
      "learning_rate": 1.041861277497315e-05,
      "loss": 0.4606,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8464,
      "eval_loss": 0.5238063931465149,
      "eval_runtime": 6.2304,
      "eval_samples_per_second": 601.89,
      "eval_steps_per_second": 9.47,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.595092296600342,
      "learning_rate": 2.89810315724514e-06,
      "loss": 0.3178,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8602666666666666,
      "eval_loss": 0.47653892636299133,
      "eval_runtime": 6.3174,
      "eval_samples_per_second": 593.595,
      "eval_steps_per_second": 9.339,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 6.809961795806885,
      "learning_rate": 1.0680510444061043e-10,
      "loss": 0.2591,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8650666666666667,
      "eval_loss": 0.46875330805778503,
      "eval_runtime": 6.3311,
      "eval_samples_per_second": 592.318,
      "eval_steps_per_second": 9.319,
      "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
}