dataset.py

# dataset.py (place at the repo root)
import os, glob, datasets

_CITATION = ""
_DESCRIPTION = "ATLAS-1: mesh dataset of single-cell 3D OBJ files."
_HOMEPAGE = "https://huggingface.co/datasets/Sentinal4D/ATLAS-1"
_LICENSE = "CC-BY-4.0"  # change if needed

class AtlasConfig(datasets.BuilderConfig):
    def __init__(self, **kwargs):
        super().__init__(version=datasets.Version("1.0.0"), **kwargs)

class Atlas(datasets.GeneratorBasedBuilder):
    BUILDER_CONFIGS = [AtlasConfig(name="default", description=_DESCRIPTION)]

    def _info(self):
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=datasets.Features({
                "id": datasets.Value("string"),
                "path": datasets.Value("string"),      # local cached path to the .obj
                # Add whatever metadata you want here:
                # "drug": datasets.Value("string"),
                # "cell_line": datasets.Value("string"),
            }),
            homepage=_HOMEPAGE,
            license=_LICENSE,
            citation=_CITATION,
        )

    def _split_generators(self, dl_manager):
        # Let users override with data_files=... if they want
        data_files = self.config.data_files or {"train": ["data/**/*.obj"]}
        # This resolves HF repo-relative patterns, downloads/caches them
        downloaded = dl_manager.download(data_files)
        return [
            datasets.SplitGenerator(name=getattr(datasets.Split, split.upper(), datasets.Split.TRAIN),
                                    gen_kwargs={"files": downloaded[split]})
            for split in downloaded
        ]

    def _generate_examples(self, files):
        # files is a list of local cached .obj paths
        for idx, p in enumerate(sorted(files)):
            yield idx, {
                "id": os.path.splitext(os.path.basename(p))[0],
                "path": p,
            }