ChatTube

ChatTube is a web application that allows users to interact with YouTube videos through a question-answer interface powered by large language models (LLMs). The app accepts a YouTube video URL, downloads the video, transcribes it using OpenAI Whisper, and employs Retrieval-Augmented Generation (RAG) to provide contextual answers based on the video content.

Features

• Video Transcription: Automatically transcribes YouTube videos using OpenAI Whisper.
• Contextual Q&A: Fetches relevant context from a vector database to generate accurate answers to user questions.
• LLM-Powered Responses: Uses Ollama LLM with the retrieved context to provide intelligent responses.
• Fast and Intuitive Interface: Built with a ReactJS frontend and FASTAPI backend for seamless interaction.

Tech Stack

• Backend: FASTAPI - Provides API endpoints for video processing, transcription, and LLM interactions.
• Frontend: ReactJS - User-friendly interface to input YouTube URLs and ask questions.
• Transcription: OpenAI Whisper - High-accuracy transcription of video audio.
• Question Answering: RAG (Retrieval-Augmented Generation) - Enhances LLM responses by providing relevant context from a vector database.
• Language Model: Ollama LLM - Processes the context and generates human-like answers.

How It Works

1. Input a YouTube URL: The user provides the URL of a YouTube video.
2. Video Processing: The backend downloads the video and transcribes the audio using Whisper.
3. Context Retrieval: The transcription is stored in a vector database, and relevant parts of the text are fetched using RAG based on the user’s query.
4. Question Answering: The retrieved context is passed to Ollama LLM, which generates a response to the user’s question.
5. Display: The answer is displayed in the frontend for the user.

Setup Instructions

Backend (FASTAPI)

1. Clone the repository:

   git clone https://github.com/reesshhoo/ChatTube.git
   cd ChatTube

2. Install dependencies:

pip install -r requirements.txt

3. Start the FASTAPI server:

uvicorn main:app --reload

Frontend (ReactJS)

1. Navigate to the frontend directory:

cd frontend

2. Install frontend dependencies:

npm install

3.Start the React development server:

npm start