download_coco.py

import os
import skimage.io as io
import pandas as pd

for split in ["train", "val", "test"]:
    os.makedirs(split, exist_ok=True)
    df = pd.read_parquet(f"{split}_images.parquet")
    for idx, row in df.iterrows():
        url = row['coco_url']
        filename = url.split('/')[-1]   
        I = io.imread(url)
        io.imsave(os.path.join(split, filename), I)