search

mlsanigeria / speak-to-docs

Speech-enabled Retrieval-Augmented Generation solution
https://speak-to-docs.streamlit.app/
MIT License
10 stars 31 forks source link

Implement Conversation History-Aware RAG Solution 🚀 #5

Open Odeyiany2 opened 2 months ago

Odeyiany2 commented 2 months ago

Implement Conversation History-Aware RAG Solution 🚀

Project Overview

We are looking to enhance our RAG (Retrieval-Augmented Generation) system with conversation history awareness. The goal is to allow the system to properly handle, and answer follow-up questions based on previous interactions, ensuring a seamless flow of information.

What is Needed

The task is to implement a conversation history-aware RAG solution that tracks and remembers previous conversations, allowing it to provide context-aware answers to follow-up questions. Efficiently manages conversation history to avoid performance issues, especially when handling long conversations.

How to Contribute

  1. Get Familiar: Review the existing RAG system, focusing on how queries are currently processed and how session management (if any) is implemented.
  2. Implement: Add functionality to track conversation history and ensure the RAG system can reference this history when responding to follow-up questions.
  3. Test Your Code: Ensure that conversation history is preserved across queries and that the RAG solution provides accurate and context-aware answers.
  4. Submit a Pull Request: Once your implementation is complete, submit a pull request with your changes, along with relevant documentation explaining your solution.

Getting Started

Before you begin, ensure that you have read through the Contribution Guidelines and have a good understanding of the RAG system architecture.

We are looking forward to your contributions! Happy Hacking! 🎉

MbuviM commented 1 month ago

Hey, kindly assign me this.