trainer_state.json

{
  "best_metric": 0.8738666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0015/checkpoints/checkpoint-1998",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1998,
  "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": 2.1144,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7754666666666666,
      "eval_loss": 0.907692551612854,
      "eval_runtime": 6.326,
      "eval_samples_per_second": 592.795,
      "eval_steps_per_second": 9.327,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 5e-05,
      "loss": 0.6655,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8296,
      "eval_loss": 0.6175572872161865,
      "eval_runtime": 6.3136,
      "eval_samples_per_second": 593.956,
      "eval_steps_per_second": 9.345,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.0579838752746582,
      "learning_rate": 5e-05,
      "loss": 0.419,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8424,
      "eval_loss": 0.5456655025482178,
      "eval_runtime": 6.3579,
      "eval_samples_per_second": 589.822,
      "eval_steps_per_second": 9.28,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.124746799468994,
      "learning_rate": 5e-05,
      "loss": 0.2978,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8538666666666667,
      "eval_loss": 0.531421422958374,
      "eval_runtime": 6.3147,
      "eval_samples_per_second": 593.85,
      "eval_steps_per_second": 9.343,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.4101334810256958,
      "learning_rate": 5e-05,
      "loss": 0.2074,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8626666666666667,
      "eval_loss": 0.5012107491493225,
      "eval_runtime": 6.6705,
      "eval_samples_per_second": 562.177,
      "eval_steps_per_second": 8.845,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 17.343618392944336,
      "learning_rate": 5e-05,
      "loss": 0.1595,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8738666666666667,
      "eval_loss": 0.46948665380477905,
      "eval_runtime": 6.45,
      "eval_samples_per_second": 581.392,
      "eval_steps_per_second": 9.147,
      "step": 1998
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.88447923735552e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}