trainer_state.json

{
  "best_metric": 0.8725333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0662/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": 3e-05,
      "loss": 2.328,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7325333333333334,
      "eval_loss": 1.1330370903015137,
      "eval_runtime": 11.3375,
      "eval_samples_per_second": 330.761,
      "eval_steps_per_second": 5.204,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.7274011373519897,
      "learning_rate": 3e-05,
      "loss": 0.7495,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8141333333333334,
      "eval_loss": 0.6692166924476624,
      "eval_runtime": 11.5331,
      "eval_samples_per_second": 325.152,
      "eval_steps_per_second": 5.116,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.5429956912994385,
      "learning_rate": 3e-05,
      "loss": 0.4187,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8413333333333334,
      "eval_loss": 0.527319610118866,
      "eval_runtime": 11.1969,
      "eval_samples_per_second": 334.915,
      "eval_steps_per_second": 5.269,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.355974793434143,
      "learning_rate": 3e-05,
      "loss": 0.2703,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8533333333333334,
      "eval_loss": 0.48818737268447876,
      "eval_runtime": 11.47,
      "eval_samples_per_second": 326.939,
      "eval_steps_per_second": 5.144,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.686638593673706,
      "learning_rate": 3e-05,
      "loss": 0.182,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8533333333333334,
      "eval_loss": 0.5072634816169739,
      "eval_runtime": 11.2303,
      "eval_samples_per_second": 333.917,
      "eval_steps_per_second": 5.254,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 18.80087661743164,
      "learning_rate": 3e-05,
      "loss": 0.1277,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8661333333333333,
      "eval_loss": 0.4705181419849396,
      "eval_runtime": 11.4418,
      "eval_samples_per_second": 327.747,
      "eval_steps_per_second": 5.157,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.082244873046875,
      "learning_rate": 3e-05,
      "loss": 0.0861,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8725333333333334,
      "eval_loss": 0.46933284401893616,
      "eval_runtime": 11.6413,
      "eval_samples_per_second": 322.13,
      "eval_steps_per_second": 5.068,
      "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
}