upload_to_hf.py

#!/usr/bin/env python3
"""
Upload AI Tool Pool Jewelry Vision dataset to Hugging Face Hub
"""
import os
from huggingface_hub import HfApi, create_repo
from pathlib import Path

def upload_jewelry_dataset():
    """Upload the jewelry vision dataset to Hugging Face Hub"""
    
    # Dataset configuration
    dataset_name = "ai-tool-pool-jewelry-vision"  # Change this to your desired repo name
    # Replace YOUR_USERNAME with your HF username
    repo_id = f"bzcasper/{dataset_name}"
    
    # Get current directory (dataset root)
    dataset_path = Path(__file__).parent
    
    print(f"Dataset path: {dataset_path}")
    print(f"Repository ID: {repo_id}")
    
    # Initialize HF API
    api = HfApi()
    
    try:
        # Create repository on Hugging Face Hub
        print("Creating repository on Hugging Face Hub...")
        create_repo(
            repo_id=repo_id,
            repo_type="dataset",
            exist_ok=True,
            private=False  # Set to True if you want private repo
        )
        print(f"✅ Repository created: https://huggingface.co/datasets/{repo_id}")
        
        # Upload all files to the repository
        print("Uploading dataset files...")
        
        # Upload the README first
        api.upload_file(
            path_or_fileobj=dataset_path / "README.md",
            path_in_repo="README.md",
            repo_id=repo_id,
            repo_type="dataset",
        )
        print("✅ README.md uploaded")
        
        # Upload dataset files by folder
        for split_dir in ["train", "test", "valid"]:
            split_path = dataset_path / split_dir
            if split_path.exists():
                print(f"Uploading {split_dir} split...")
                api.upload_folder(
                    folder_path=split_path,
                    path_in_repo=split_dir,
                    repo_id=repo_id,
                    repo_type="dataset",
                    ignore_patterns=["*.txt"],  # Skip tokenization files
                )
                print(f"✅ {split_dir} split uploaded")
        
        # Upload original README files for reference
        for readme_file in ["README.dataset.txt", "README.roboflow.txt"]:
            readme_path = dataset_path / readme_file
            if readme_path.exists():
                api.upload_file(
                    path_or_fileobj=readme_path,
                    path_in_repo=readme_file,
                    repo_id=repo_id,
                    repo_type="dataset",
                )
                print(f"✅ {readme_file} uploaded")
        
        print(f"\n🎉 Dataset successfully uploaded!")
        print(f"📋 Dataset URL: https://huggingface.co/datasets/{repo_id}")
        print(f"📁 Total files uploaded from train, test, and valid splits")
        
    except Exception as e:
        print(f"❌ Error uploading dataset: {str(e)}")
        print("\nMake sure you:")
        print("1. Are logged in: huggingface-cli login")
        print("2. Have updated the repo_id with your username")
        print("3. Have write permissions to the repository")

if __name__ == "__main__":
    upload_jewelry_dataset()