photos.py

import os
import zipfile
from pathlib import Path
import datasets

class Photos(datasets.GeneratorBasedBuilder):
    VERSION = datasets.Version("1.0.0")

    def _info(self):
        return datasets.DatasetInfo(
            features=datasets.Features({
                "image": datasets.Image(),
                "label": datasets.ClassLabel(names=["Not Applicable", "Very Poor", "Poor", "Fair", "Good", "Excellent", "Exceptional"]),
            }),
            supervised_keys=("image", "label"),
        )

    def _split_generators(self, dl_manager):
        # Define the URLs for the zip files
        urls = {
            'Not Applicable': "https://huggingface.co/datasets/rshrott/photos/resolve/main/Not Applicable.zip",
            'Very Poor': "https://huggingface.co/datasets/rshrott/photos/resolve/main/Very Poor.zip",
            'Poor': "https://huggingface.co/datasets/rshrott/photos/resolve/main/Poor.zip",
            'Fair': "https://huggingface.co/datasets/rshrott/photos/resolve/main/Fair.zip",
            'Good': "https://huggingface.co/datasets/rshrott/photos/resolve/main/Good.zip",
            'Excellent': "https://huggingface.co/datasets/rshrott/photos/resolve/main/Excellent.zip",
            'Exceptional': "https://huggingface.co/datasets/rshrott/photos/resolve/main/Exceptional.zip"
        }

        # Download and extract the zip files
        downloaded_files = dl_manager.download_and_extract(urls)
        extracted_dirs = {label: Path(file).stem for label, file in downloaded_files.items()}  # Remove .zip extension

        # Here you would split the dataset into train, validation, and test sets
        # For simplicity, we'll assume all images are used for training
        return [
            datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"extracted_dirs": extracted_dirs}),
        ]

    def _generate_examples(self, extracted_dirs):
        # Iterate over the images in the extracted directories and yield examples
        for label, dir in extracted_dirs.items():
            label_dir = os.path.join(self.config.data_dir, dir)
            for img_path in Path(label_dir).glob('*.jpeg'):
                yield str(img_path), {
                    "image": str(img_path),
                    "label": label,
                }