trainer_state.json

{
  "best_metric": 0.8656,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0688/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": 3e-05,
      "loss": 2.2792,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7397333333333334,
      "eval_loss": 1.1496541500091553,
      "eval_runtime": 6.1411,
      "eval_samples_per_second": 610.636,
      "eval_steps_per_second": 9.607,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 3e-05,
      "loss": 0.7844,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8173333333333334,
      "eval_loss": 0.7127450108528137,
      "eval_runtime": 6.0968,
      "eval_samples_per_second": 615.074,
      "eval_steps_per_second": 9.677,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 3e-05,
      "loss": 0.4447,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8469333333333333,
      "eval_loss": 0.5403169989585876,
      "eval_runtime": 6.0871,
      "eval_samples_per_second": 616.061,
      "eval_steps_per_second": 9.693,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 79.46273803710938,
      "learning_rate": 3e-05,
      "loss": 0.2973,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8530666666666666,
      "eval_loss": 0.5166841149330139,
      "eval_runtime": 6.0608,
      "eval_samples_per_second": 618.735,
      "eval_steps_per_second": 9.735,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 22.246057510375977,
      "learning_rate": 3e-05,
      "loss": 0.199,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8656,
      "eval_loss": 0.4715491533279419,
      "eval_runtime": 6.6467,
      "eval_samples_per_second": 564.189,
      "eval_steps_per_second": 8.877,
      "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
}