# Description: Preload the dataset to cache_dir
# Copyright (C) 2024 Ronan Le Meillat
# License: Apache License 2.0
from datasets import load_dataset
dataset_id = "eltorio/ROCO-radiology"
cache_dir = "/workspace/data"
train_dataset = load_dataset(dataset_id, split="train", cache_dir=cache_dir)