trainer_state.json

{
  "best_metric": 0.8546666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0524/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": 9e-05,
      "loss": 1.5919,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7792,
      "eval_loss": 0.7710332870483398,
      "eval_runtime": 11.0347,
      "eval_samples_per_second": 339.838,
      "eval_steps_per_second": 5.347,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6714239120483398,
      "learning_rate": 9e-05,
      "loss": 0.5632,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.812,
      "eval_loss": 0.6055018305778503,
      "eval_runtime": 11.3636,
      "eval_samples_per_second": 330.001,
      "eval_steps_per_second": 5.192,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.9081592559814453,
      "learning_rate": 9e-05,
      "loss": 0.3796,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8426666666666667,
      "eval_loss": 0.5128757953643799,
      "eval_runtime": 11.7011,
      "eval_samples_per_second": 320.483,
      "eval_steps_per_second": 5.042,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 9e-05,
      "loss": 0.2754,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8525333333333334,
      "eval_loss": 0.505230188369751,
      "eval_runtime": 12.8871,
      "eval_samples_per_second": 290.988,
      "eval_steps_per_second": 4.578,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.35341235995292664,
      "learning_rate": 9e-05,
      "loss": 0.2036,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8546666666666667,
      "eval_loss": 0.5069410800933838,
      "eval_runtime": 11.1206,
      "eval_samples_per_second": 337.213,
      "eval_steps_per_second": 5.305,
      "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
}