trainer_state.json

{
  "best_metric": 0.8746666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0602/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": 5e-05,
      "loss": 1.9944,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8024,
      "eval_loss": 0.8258088827133179,
      "eval_runtime": 6.7437,
      "eval_samples_per_second": 556.072,
      "eval_steps_per_second": 8.749,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 5e-05,
      "loss": 0.5786,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8392,
      "eval_loss": 0.5915732979774475,
      "eval_runtime": 6.2104,
      "eval_samples_per_second": 603.825,
      "eval_steps_per_second": 9.5,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.7018661499023438,
      "learning_rate": 5e-05,
      "loss": 0.3485,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8626666666666667,
      "eval_loss": 0.4830835163593292,
      "eval_runtime": 7.2281,
      "eval_samples_per_second": 518.81,
      "eval_steps_per_second": 8.163,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 91.14945983886719,
      "learning_rate": 5e-05,
      "loss": 0.2438,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8698666666666667,
      "eval_loss": 0.4540748596191406,
      "eval_runtime": 6.2534,
      "eval_samples_per_second": 599.67,
      "eval_steps_per_second": 9.435,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.1823046207427979,
      "learning_rate": 5e-05,
      "loss": 0.1735,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8746666666666667,
      "eval_loss": 0.45350098609924316,
      "eval_runtime": 6.033,
      "eval_samples_per_second": 621.586,
      "eval_steps_per_second": 9.78,
      "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
}