upload_dataset.py

#!/usr/bin/env python3
"""
Script to upload the bimanual bone dataset to Hugging Face
"""

import os
import zipfile
import tempfile
from pathlib import Path
from huggingface_hub import HfApi, create_repo
import argparse

def extract_and_upload_dataset(zip_path, repo_id, folder_name="so101"):
    """
    Extract the zip file and upload to Hugging Face dataset repository
    """
    api = HfApi()
    
    # Create the repository if it doesn't exist
    try:
        create_repo(repo_id=repo_id, repo_type="dataset", exist_ok=True)
        print(f"Repository {repo_id} is ready")
    except Exception as e:
        print(f"Repository creation/verification: {e}")
    
    # Create temporary directory for extraction
    with tempfile.TemporaryDirectory() as temp_dir:
        print(f"Extracting {zip_path} to temporary directory...")
        
        # Extract the zip file
        with zipfile.ZipFile(zip_path, 'r') as zip_ref:
            zip_ref.extractall(temp_dir)
        
        # Find the extracted content
        extracted_items = list(Path(temp_dir).iterdir())
        if len(extracted_items) == 1 and extracted_items[0].is_dir():
            # If there's a single directory, use its contents
            source_dir = extracted_items[0]
        else:
            # Otherwise use the temp directory itself
            source_dir = Path(temp_dir)
        
        # Create the target folder structure
        target_dir = source_dir / folder_name
        target_dir.mkdir(exist_ok=True)
        
        # Move all extracted content to the so101 folder
        for item in source_dir.iterdir():
            if item.name != folder_name:
                import shutil
                shutil.move(str(item), str(target_dir / item.name))
        
        print(f"Uploading dataset to {repo_id}...")
        print(f"Source directory: {source_dir}")
        print(f"Target folder: {folder_name}")
        
        # Upload the dataset with pull request
        api.upload_folder(
            folder_path=str(source_dir),
            repo_id=repo_id,
            repo_type="dataset",
            commit_message="Upload bimanual bone packing dataset with so101 folder structure",
            create_pr=True
        )
        
        print("Dataset uploaded successfully!")

def main():
    parser = argparse.ArgumentParser(description="Upload bimanual bone dataset to Hugging Face")
    parser.add_argument("--zip-path", required=True, help="Path to the zip file")
    parser.add_argument("--repo-id", default="CHEWYSO/Chewy_Robotics_Bone_Bi-manual_Packing", 
                       help="Hugging Face repository ID")
    parser.add_argument("--folder-name", default="so101", 
                       help="Folder name to organize the dataset")
    
    args = parser.parse_args()
    
    if not os.path.exists(args.zip_path):
        print(f"Error: Zip file {args.zip_path} not found")
        return 1
    
    extract_and_upload_dataset(args.zip_path, args.repo_id, args.folder_name)
    return 0

if __name__ == "__main__":
    exit(main())