upload_openclip.py

"""Upload exported OpenCLIP .pte files and metadata to HuggingFace Hub."""

import shutil
from pathlib import Path

from huggingface_hub import HfApi

REPO_ID = "mallman/openclip-vit-l14-coreml"
HF_REPO_DIR = Path("hf_repo_openclip")
OUTPUT_DIR = Path("output/openclip")


def main():
    api = HfApi()

    # Create the repo (no-op if it already exists)
    api.create_repo(REPO_ID, repo_type="model", exist_ok=True)
    print(f"Repo: https://huggingface.co/{REPO_ID}")

    HF_REPO_DIR.mkdir(exist_ok=True)

    # Copy config
    config_src = OUTPUT_DIR / "config.json"
    if config_src.exists():
        shutil.copy(config_src, HF_REPO_DIR / "config.json")

    # Copy .pte files
    for pte in OUTPUT_DIR.glob("clip_vit_l14_*.pte"):
        dest = HF_REPO_DIR / pte.name
        if not dest.exists() or dest.stat().st_mtime < pte.stat().st_mtime:
            print(f"Copying {pte.name} to staging dir...")
            shutil.copy(pte, dest)

    # Copy tokenizer files
    for fname in ["vocab.json", "merges.txt"]:
        src = OUTPUT_DIR / fname
        if src.exists():
            shutil.copy(src, HF_REPO_DIR / fname)

    # Upload everything
    print("Uploading to HuggingFace Hub...")
    api.upload_folder(
        folder_path=str(HF_REPO_DIR),
        repo_id=REPO_ID,
        repo_type="model",
    )
    print(f"Done! https://huggingface.co/{REPO_ID}")


if __name__ == "__main__":
    main()