trainer_state.json

{
  "best_metric": 0.928,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0050/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": 0.17561949789524078,
      "learning_rate": 0.0001,
      "loss": 0.6049,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8992,
      "eval_loss": 0.33713605999946594,
      "eval_runtime": 14.6968,
      "eval_samples_per_second": 255.157,
      "eval_steps_per_second": 4.014,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.05850747600197792,
      "learning_rate": 0.0001,
      "loss": 0.1432,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9242666666666667,
      "eval_loss": 0.24559389054775238,
      "eval_runtime": 13.2542,
      "eval_samples_per_second": 282.929,
      "eval_steps_per_second": 4.451,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.023324014618992805,
      "learning_rate": 0.0001,
      "loss": 0.0846,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9170666666666667,
      "eval_loss": 0.3136230409145355,
      "eval_runtime": 12.2799,
      "eval_samples_per_second": 305.377,
      "eval_steps_per_second": 4.805,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.020313702523708344,
      "learning_rate": 0.0001,
      "loss": 0.0597,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9216,
      "eval_loss": 0.3027634620666504,
      "eval_runtime": 11.8737,
      "eval_samples_per_second": 315.824,
      "eval_steps_per_second": 4.969,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.011006979271769524,
      "learning_rate": 0.0001,
      "loss": 0.0449,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9189333333333334,
      "eval_loss": 0.34608912467956543,
      "eval_runtime": 12.0901,
      "eval_samples_per_second": 310.17,
      "eval_steps_per_second": 4.88,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.007892117835581303,
      "learning_rate": 0.0001,
      "loss": 0.0377,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9224,
      "eval_loss": 0.3170318007469177,
      "eval_runtime": 12.0473,
      "eval_samples_per_second": 311.273,
      "eval_steps_per_second": 4.897,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 0.0001,
      "loss": 0.0345,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.928,
      "eval_loss": 0.31575509905815125,
      "eval_runtime": 12.3191,
      "eval_samples_per_second": 304.405,
      "eval_steps_per_second": 4.789,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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
}