trainer_state.json

{
  "best_metric": 0.8944,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0166/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "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": 6.655434848731736e-05,
      "loss": 1.8282,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7709333333333334,
      "eval_loss": 0.8897823095321655,
      "eval_runtime": 6.6042,
      "eval_samples_per_second": 567.822,
      "eval_steps_per_second": 8.934,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 5.689582334051816e-05,
      "loss": 0.6168,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8269333333333333,
      "eval_loss": 0.5937406420707703,
      "eval_runtime": 6.5777,
      "eval_samples_per_second": 570.108,
      "eval_steps_per_second": 8.97,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 30.200910568237305,
      "learning_rate": 4.288018105154487e-05,
      "loss": 0.3787,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8578666666666667,
      "eval_loss": 0.5036105513572693,
      "eval_runtime": 6.3603,
      "eval_samples_per_second": 589.595,
      "eval_steps_per_second": 9.276,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 76.6411361694336,
      "learning_rate": 2.7303772261234784e-05,
      "loss": 0.2272,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8722666666666666,
      "eval_loss": 0.4221237003803253,
      "eval_runtime": 5.8166,
      "eval_samples_per_second": 644.704,
      "eval_steps_per_second": 10.143,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 21.74244499206543,
      "learning_rate": 1.325169576262615e-05,
      "loss": 0.131,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8842666666666666,
      "eval_loss": 0.4006495475769043,
      "eval_runtime": 5.9615,
      "eval_samples_per_second": 629.041,
      "eval_steps_per_second": 9.897,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.5619018077850342,
      "learning_rate": 3.5071376436769933e-06,
      "loss": 0.0729,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8914666666666666,
      "eval_loss": 0.3678758442401886,
      "eval_runtime": 5.7536,
      "eval_samples_per_second": 651.763,
      "eval_steps_per_second": 10.254,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.2558310031890869,
      "learning_rate": 1.2714900776311653e-10,
      "loss": 0.0495,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8944,
      "eval_loss": 0.37385794520378113,
      "eval_runtime": 6.1886,
      "eval_samples_per_second": 605.957,
      "eval_steps_per_second": 9.534,
      "step": 2331
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}