Overview

This project is a YouTube Video Summarization App that leverages LangChain, Llama3.2, YouTube Transcript API, and Streamlit to provide concise summaries of YouTube video transcripts. The app extracts the transcript from a given YouTube video URL and generates a summary using a local language model.

Features

• Extracts transcripts from YouTube videos
• Summarizes transcripts using a local LLM (Llama3.2, Mistral, or CodeLlama)
• Adjustable summary length
• User-friendly interface with Streamlit
• Options to download the transcript and summary

Installation

1. Clone the repository:

   git clone https://github.com/yourusername/YouTube-Video-Summarization-App.git
   cd YouTube-Video-Summarization-App

2. Install the required dependencies:

   pip install -r requirements.txt

Usage

1. Run the Streamlit app:

   streamlit run app.py

2. Open your web browser and go to http://localhost:8501.

3. Enter the YouTube video URL and adjust the settings as needed.

4. Click the button to generate the summary and view the results.

Configuration

• Model Selection: Choose between Llama3.2, Mistral, or CodeLlama for summarization.
• Summary Length: Adjust the length of the generated summary using the slider in the sidebar.

Contributing

Contributions are welcome! Please fork the repository and submit a pull request.

License

This project is licensed under the MIT License. See the LICENSE file for details.

Acknowledgements

• Streamlit
• YouTube Transcript API
• LangChain
• Meta LLaMA3