https://github.com/farhaj499/multi_ai_agent_with_rag_and_wikisearch

https://github.com/farhaj499/multi_ai_agent_with_rag_and_wikisearch

Last synced: 1 day ago
JSON representation

• Host: GitHub
• URL: https://github.com/farhaj499/multi_ai_agent_with_rag_and_wikisearch
• Owner: Farhaj499
• License: mit
• Created: 2024-12-30T12:23:50.000Z (about 1 month ago)
• Default Branch: main
• Last Pushed: 2025-01-12T17:36:22.000Z (18 days ago)
• Last Synced: 2025-01-12T18:31:38.261Z (18 days ago)
• Language: Jupyter Notebook
• Size: 33.2 KB
• Stars: 1
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Multi AI Agent with RAG and WikiSearch

This project demonstrates the integration of multiple AI agents using Retrieval-Augmented Generation (RAG) and WikiSearch functionalities, along with Cassandra for data storage and retrieval. It leverages Llama 3.1 for natural language processing and "all-MiniLM-L6-v2" for generating embeddings.

## Overview

This project, implemented in a Jupyter Notebook, showcases how to:

*   Connect to Cassandra.

*   Load and process data from web pages using `WebBaseLoader`.

*   Convert text to vectors using the "all-MiniLM-L6-v2" Hugging Face embedding model.

*   Perform document retrieval and question routing using LangChain and related libraries.

*   Generate human-like responses using the Llama 3.1 LLM via GROQ.

## Features

*   **Cassandra Integration:** Connects to Cassandra for storing and retrieving document embeddings.

*   **LangChain:** Utilizes LangChain for web page loading, text splitting, embedding, and vector store operations.

*   **WebBaseLoader:** Uses `WebBaseLoader` to efficiently load content from web pages.

*   **Hugging Face Embeddings:** Employs the "all-MiniLM-L6-v2" model for generating high-quality text embeddings.

*   **Llama 3.1 with GROQ:** Integrates the Llama 3.1 LLM using GROQ for refined, human-like responses.

*   **WikiSearch:** Integrates the Wikipedia API for answering questions not related to stored documents.

*   **Query Routing:** Routes user queries to either the Cassandra vector store or WikiSearch based on relevance.