DEPLOYMENT.md

# 🚀 Deployment Guide for Hugging Face Spaces

## Files Created for Deployment

1. **app.py** - Main Gradio application
2. **README.md** - Space configuration and documentation
3. **requirements.txt** - Python dependencies
4. **packages.txt** - System dependencies for Manim
5. **.gitignore** - Files to exclude from git
6. **test_deployment.py** - Local testing script

## Step-by-Step Deployment Instructions

### 1. Create a New Hugging Face Space

1. Go to [Hugging Face Spaces](https://huggingface.co/spaces)
2. Click "Create new Space"
3. Choose:
   - **Space name**: `ai-video-generator` (or your preferred name)
   - **License**: MIT
   - **SDK**: Gradio
   - **Hardware**: CPU Basic (free tier) or GPU if you have Pro subscription

### 2. Upload Files

Upload all the files from your video_generator folder:
- `app.py`
- `README.md`
- `requirements.txt`
- `packages.txt`
- `.gitignore`

**DO NOT upload**:
- `.env` file (contains your API key)
- `video_generator.py` (original file, not needed)

### 3. Set Environment Variables

1. Go to your Space settings
2. Click on "Variables and secrets"
3. Add a new secret:
   - **Name**: `GEMINI_API_KEY`
   - **Value**: Your actual Gemini API key (from your .env file)

### 4. Wait for Build

- The space will automatically build and install dependencies
- This may take 5-15 minutes due to Manim installation
- Check the build logs for any errors

### 5. Test the Application

1. Once built, test with a simple topic like "Introduction to Python"
2. Check if videos generate correctly
3. Verify the interface works smoothly

## Troubleshooting

### Common Issues and Solutions

1. **Build Timeout**
   - Manim installation can be slow
   - Try using a smaller requirements.txt with fewer versions specified

2. **Memory Issues**
   - Use shorter video durations (1 minute)
   - Consider upgrading to GPU hardware

3. **API Key Issues**
   - Double-check the secret name is exactly `GEMINI_API_KEY`
   - Verify the API key is valid and has quota

4. **Video Generation Fails**
   - Check the generation logs in the interface
   - Manim syntax might need adjustment for the cloud environment

## Alternative Deployment (GitHub → HF Spaces)

If you prefer using Git:

1. Create a GitHub repository
2. Push all files (except .env)
3. Connect the repository to Hugging Face Spaces
4. Set the API key in Spaces secrets

## Performance Tips

1. **Use shorter durations** (1-2 minutes max)
2. **Simple topics** work better than complex ones
3. **Monitor usage** to stay within API limits
4. **Consider caching** for frequently requested topics

## Security Notes

- Never commit API keys to public repositories
- Use Hugging Face Spaces secrets for sensitive data
- The .gitignore file excludes sensitive files

## Support

If you encounter issues:
1. Check the Space logs
2. Verify all dependencies in requirements.txt
3. Test locally first with test_deployment.py
4. Consider the Hugging Face Spaces documentation