trainer_state.json

{
  "best_metric": 0.8730666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0055/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": 2.0736894607543945,
      "learning_rate": 0.0003,
      "loss": 0.6125,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8466666666666667,
      "eval_loss": 0.5239955186843872,
      "eval_runtime": 10.9149,
      "eval_samples_per_second": 343.567,
      "eval_steps_per_second": 5.405,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.32727551460266113,
      "learning_rate": 0.0003,
      "loss": 0.301,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8690666666666667,
      "eval_loss": 0.46166443824768066,
      "eval_runtime": 10.9161,
      "eval_samples_per_second": 343.528,
      "eval_steps_per_second": 5.405,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.372037410736084,
      "learning_rate": 0.0003,
      "loss": 0.2127,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8722666666666666,
      "eval_loss": 0.47941914200782776,
      "eval_runtime": 11.0612,
      "eval_samples_per_second": 339.022,
      "eval_steps_per_second": 5.334,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.9062600135803223,
      "learning_rate": 0.0003,
      "loss": 0.1679,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8605333333333334,
      "eval_loss": 0.5270391702651978,
      "eval_runtime": 10.7132,
      "eval_samples_per_second": 350.037,
      "eval_steps_per_second": 5.507,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.25479844212532043,
      "learning_rate": 0.0003,
      "loss": 0.1421,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8626666666666667,
      "eval_loss": 0.528075635433197,
      "eval_runtime": 11.1185,
      "eval_samples_per_second": 337.275,
      "eval_steps_per_second": 5.306,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.16481293737888336,
      "learning_rate": 0.0003,
      "loss": 0.1172,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8690666666666667,
      "eval_loss": 0.525187611579895,
      "eval_runtime": 11.3239,
      "eval_samples_per_second": 331.157,
      "eval_steps_per_second": 5.21,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 0.0003,
      "loss": 0.1115,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8730666666666667,
      "eval_loss": 0.537767231464386,
      "eval_runtime": 10.7333,
      "eval_samples_per_second": 349.381,
      "eval_steps_per_second": 5.497,
      "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
}