Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/oussemabenhassena5/notegen-with-llama-and-whisper

AI-powered YouTube video notes generator
https://github.com/oussemabenhassena5/notegen-with-llama-and-whisper

ai llama3 python whisper youtube-api

Last synced: 6 days ago
JSON representation

AI-powered YouTube video notes generator

Host: GitHub
URL: https://github.com/oussemabenhassena5/notegen-with-llama-and-whisper
Owner: oussemabenhassena5
License: mit
Created: 2025-01-24T15:53:30.000Z (19 days ago)
Default Branch: main
Last Pushed: 2025-02-06T08:30:57.000Z (6 days ago)
Last Synced: 2025-02-06T09:33:44.636Z (6 days ago)
Topics: ai, llama3, python, whisper, youtube-api
Language: Python
Homepage:
Size: 13.7 KB
Stars: 1
Watchers: 1
Forks: 0
Open Issues: 0
Metadata Files:
- Readme: README.md

Awesome Lists containing this project

README

# 🎥 NoteGen: Transform YouTube Videos into Intelligent Notes

[![Made with Python](https://img.shields.io/badge/Made%20with-Python-1f425f.svg)](https://www.python.org/)
[![Powered by LLaMA3](https://img.shields.io/badge/Powered%20by-LLaMA3--70B-blue)](https://ai.meta.com/llama/)
[![OpenAI Whisper](https://img.shields.io/badge/ASR-OpenAI%20Whisper-green)](https://github.com/openai/whisper)

**NoteGen** harnesses the power of cutting-edge AI to revolutionize how we learn from video content. By combining Meta's LLaMA3-70B (via Groq's lightning-fast inference engine) with OpenAI's Whisper, NoteGen transforms YouTube videos into comprehensive, structured notes that capture the essence of any content.

## 🚀 Key Features

### Advanced AI Integration
- **Dual-Engine Transcription**: Seamlessly switches between native YouTube transcripts and Whisper ASR for maximum accuracy
- **State-of-the-Art Summarization**: Leverages LLaMA3-70B through Groq's ultra-efficient API for nuanced content understanding
- **Multi-Language Support**: Process content in 5 different languages with native-quality results

### User Experience
- **Interactive Video Preview**: Rich thumbnail integration for content verification
- **Conversation History**: Track and review your note-generation history
- **Streamlined Interface**: Built with Streamlit for a smooth, intuitive user experience

## 🛠️ Technical Architecture

```mermaid
graph LR
A[YouTube Video] --> B[Transcript Extraction]
B --> C[LLaMA3 Processing]
C --> D[Note Generation]
B --> |Fallback| E[Whisper ASR]
E --> C
```

## 🔧 Installation

1. **Clone & Navigate**:
```bash
git clone https://github.com/oussemabenhassena5/NoteGen-with-Llama-and-whisper.git
cd NoteGen
```

2. **Set Up Environment**:
```bash
python -m venv venv
source venv/bin/activate # Windows: `venv\Scripts\activate`
pip install -r requirements.txt
```

3. **Configure API Keys**:
```bash
# .env file
GROQ_API_KEY=your_groq_api_key_here
```

4. **Launch Application**:
```bash
streamlit run app.py
```

## 💡 Usage Example

```python
from notegen import VideoProcessor

# Initialize processor
processor = VideoProcessor(video_url="https://youtube.com/watch?v=example")

# Generate notes
notes = processor.generate_notes(
language="en",
format="detailed"
)

# Access structured content
print(notes.summary)
print(notes.key_points)
```

## 🔬 Technical Deep Dive

### AI Model Integration
- **LLaMA3-70B**: Leveraging one of the most advanced open-source language models for precise content understanding
- **Whisper ASR**: State-of-the-art speech recognition with robust noise handling
- **Custom Prompting Engine**: Sophisticated prompt engineering for optimal AI performance

### Performance Optimization
- Efficient token management for processing long videos
- Smart caching system for rapid repeated access
- Parallel processing for simultaneous video and transcript handling

## 🤝 Contributing

Your expertise can help make NoteGen even better! Here's how to contribute:

1. Fork the repository
2. Create your feature branch (`git checkout -b feature/AmazingFeature`)
3. Commit your changes (`git commit -m 'Add some AmazingFeature'`)
4. Push to the branch (`git push origin feature/AmazingFeature`)
5. Open a Pull Request

## 📚 Tech Stack

- **Frontend**: Streamlit
- **AI Models**: LLaMA3-70B, OpenAI Whisper
- **APIs**: Groq Inference API, YouTube Data API
- **Core**: Python 3.9+
- **Dependencies**: PyTube, youtube-transcript-api

## 🙏 Acknowledgements

Built with gratitude using these amazing technologies:
- [Streamlit](https://streamlit.io/) - For the sleek UI
- [YouTube Transcript API](https://github.com/jdepoix/youtube-transcript-api) - For reliable transcript extraction
- [PyTube](https://github.com/pytube/pytube) - For YouTube integration
- [OpenAI Whisper](https://github.com/openai/whisper) - For powerful ASR capabilities
- [LLaMA](https://ai.meta.com/llama/) - For state-of-the-art language processing

## 📝 License

Distributed under the MIT License. See `LICENSE` for more information.

---

**Built with ❤️ by OUSSEMA**

[LinkedIn](linkedin.com/in/oussema-ben-hassena-b445122a4)