trainer_state.json

{
  "best_metric": 0.8738666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0411/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": 4.7538820348083834e-05,
      "loss": 1.9858,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7570666666666667,
      "eval_loss": 0.9245794415473938,
      "eval_runtime": 7.5611,
      "eval_samples_per_second": 495.957,
      "eval_steps_per_second": 7.803,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.1803984642028809,
      "learning_rate": 4.0613573610928476e-05,
      "loss": 0.6257,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8245333333333333,
      "eval_loss": 0.6233515739440918,
      "eval_runtime": 5.8169,
      "eval_samples_per_second": 644.674,
      "eval_steps_per_second": 10.143,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 3.0628700751103485e-05,
      "loss": 0.3507,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8493333333333334,
      "eval_loss": 0.519737958908081,
      "eval_runtime": 5.2943,
      "eval_samples_per_second": 708.31,
      "eval_steps_per_second": 11.144,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 14.800642967224121,
      "learning_rate": 1.950269447231056e-05,
      "loss": 0.1957,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8602666666666666,
      "eval_loss": 0.4767518639564514,
      "eval_runtime": 5.6724,
      "eval_samples_per_second": 661.096,
      "eval_steps_per_second": 10.401,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.22956186532974243,
      "learning_rate": 9.465496973304394e-06,
      "loss": 0.1037,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8693333333333333,
      "eval_loss": 0.4573802947998047,
      "eval_runtime": 5.8169,
      "eval_samples_per_second": 644.677,
      "eval_steps_per_second": 10.143,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.31566867232322693,
      "learning_rate": 2.5050983169121382e-06,
      "loss": 0.0555,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8730666666666667,
      "eval_loss": 0.4609951674938202,
      "eval_runtime": 5.4049,
      "eval_samples_per_second": 693.818,
      "eval_steps_per_second": 10.916,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.4947744607925415,
      "learning_rate": 9.082071983079754e-11,
      "loss": 0.0385,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8738666666666667,
      "eval_loss": 0.4557896554470062,
      "eval_runtime": 5.6866,
      "eval_samples_per_second": 659.445,
      "eval_steps_per_second": 10.375,
      "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
}