trainer_state.json

{
  "best_metric": 0.8170666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0917/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": 0.0005,
      "loss": 0.9271,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7784,
      "eval_loss": 0.7531176209449768,
      "eval_runtime": 5.7735,
      "eval_samples_per_second": 649.516,
      "eval_steps_per_second": 10.219,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.4085859954357147,
      "learning_rate": 0.0005,
      "loss": 0.5548,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7986666666666666,
      "eval_loss": 0.6766614317893982,
      "eval_runtime": 5.7971,
      "eval_samples_per_second": 646.871,
      "eval_steps_per_second": 10.177,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 12.68967342376709,
      "learning_rate": 0.0005,
      "loss": 0.4357,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8093333333333333,
      "eval_loss": 0.6335845589637756,
      "eval_runtime": 5.8197,
      "eval_samples_per_second": 644.36,
      "eval_steps_per_second": 10.138,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 21.524168014526367,
      "learning_rate": 0.0005,
      "loss": 0.3527,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8117333333333333,
      "eval_loss": 0.6645823121070862,
      "eval_runtime": 5.7976,
      "eval_samples_per_second": 646.816,
      "eval_steps_per_second": 10.177,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 23.30427360534668,
      "learning_rate": 0.0005,
      "loss": 0.3098,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8152,
      "eval_loss": 0.6732013821601868,
      "eval_runtime": 5.8603,
      "eval_samples_per_second": 639.903,
      "eval_steps_per_second": 10.068,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.9087941646575928,
      "learning_rate": 0.0005,
      "loss": 0.2685,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.816,
      "eval_loss": 0.6811097860336304,
      "eval_runtime": 6.1083,
      "eval_samples_per_second": 613.921,
      "eval_steps_per_second": 9.659,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.013002610765397549,
      "learning_rate": 0.0005,
      "loss": 0.2466,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8170666666666667,
      "eval_loss": 0.7104143500328064,
      "eval_runtime": 5.7941,
      "eval_samples_per_second": 647.216,
      "eval_steps_per_second": 10.183,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}