trainer_state.json

{
  "best_metric": 0.9333333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0450/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": 0.2441091388463974,
      "learning_rate": 0.0002571428571428571,
      "loss": 0.6177,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8664,
      "eval_loss": 0.43419429659843445,
      "eval_runtime": 13.355,
      "eval_samples_per_second": 280.793,
      "eval_steps_per_second": 4.418,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.19917137920856476,
      "learning_rate": 0.00021428571428571427,
      "loss": 0.2583,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8802666666666666,
      "eval_loss": 0.40264785289764404,
      "eval_runtime": 13.0789,
      "eval_samples_per_second": 286.722,
      "eval_steps_per_second": 4.511,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.2558488845825195,
      "learning_rate": 0.0001714285714285714,
      "loss": 0.1514,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8954666666666666,
      "eval_loss": 0.3843840956687927,
      "eval_runtime": 13.6633,
      "eval_samples_per_second": 274.458,
      "eval_steps_per_second": 4.318,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.013486538082361221,
      "learning_rate": 0.00012857142857142855,
      "loss": 0.0867,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9090666666666667,
      "eval_loss": 0.361663281917572,
      "eval_runtime": 13.7025,
      "eval_samples_per_second": 273.674,
      "eval_steps_per_second": 4.306,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.012008058838546276,
      "learning_rate": 8.57142857142857e-05,
      "loss": 0.0398,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9165333333333333,
      "eval_loss": 0.32851526141166687,
      "eval_runtime": 14.1598,
      "eval_samples_per_second": 264.834,
      "eval_steps_per_second": 4.167,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.003940851427614689,
      "learning_rate": 4.285714285714285e-05,
      "loss": 0.0171,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9309333333333333,
      "eval_loss": 0.2902136743068695,
      "eval_runtime": 13.2222,
      "eval_samples_per_second": 283.614,
      "eval_steps_per_second": 4.462,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 1.2870012870012868e-07,
      "loss": 0.0041,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9333333333333333,
      "eval_loss": 0.2786957621574402,
      "eval_runtime": 13.1197,
      "eval_samples_per_second": 285.83,
      "eval_steps_per_second": 4.497,
      "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
}