trainer_state.json

{
  "best_metric": 0.8728,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0886/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": 1.0368582010269165,
      "learning_rate": 0.0003,
      "loss": 0.6482,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8450666666666666,
      "eval_loss": 0.4905000627040863,
      "eval_runtime": 4.6196,
      "eval_samples_per_second": 811.754,
      "eval_steps_per_second": 12.772,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.98698091506958,
      "learning_rate": 0.0003,
      "loss": 0.3177,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8453333333333334,
      "eval_loss": 0.5207892656326294,
      "eval_runtime": 4.1332,
      "eval_samples_per_second": 907.295,
      "eval_steps_per_second": 14.275,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.05433693528175354,
      "learning_rate": 0.0003,
      "loss": 0.2323,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8578666666666667,
      "eval_loss": 0.514764130115509,
      "eval_runtime": 4.1639,
      "eval_samples_per_second": 900.602,
      "eval_steps_per_second": 14.169,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0003,
      "loss": 0.1905,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.856,
      "eval_loss": 0.5164419412612915,
      "eval_runtime": 4.1899,
      "eval_samples_per_second": 895.017,
      "eval_steps_per_second": 14.082,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.1076078414916992,
      "learning_rate": 0.0003,
      "loss": 0.1498,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8485333333333334,
      "eval_loss": 0.578416645526886,
      "eval_runtime": 4.2298,
      "eval_samples_per_second": 886.558,
      "eval_steps_per_second": 13.949,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.05842569097876549,
      "learning_rate": 0.0003,
      "loss": 0.1353,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8442666666666667,
      "eval_loss": 0.6676133275032043,
      "eval_runtime": 4.2151,
      "eval_samples_per_second": 889.661,
      "eval_steps_per_second": 13.997,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.28423312306404114,
      "learning_rate": 0.0003,
      "loss": 0.1204,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8728,
      "eval_loss": 0.5332277417182922,
      "eval_runtime": 4.2337,
      "eval_samples_per_second": 885.756,
      "eval_steps_per_second": 13.936,
      "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
}