RAG Chatbot with LangChain

This repository contains scripts to build and run a Retrieval-Augmented Generation (RAG) chatbot using LangChain. The project focuses on leveraging language models to provide context-aware and relevant responses, making it suitable for applications requiring information retrieval and conversational AI.

Table of Contents

• File Descriptions
  • build_rag.py
  • chatbot.py
  • ollama_llm.py
• Dependencies
• Usage
  • Running the Scripts
  • Build the RAG Pipeline
  • Run the Chatbot
• Notes

File Descriptions

build_rag.py

This script is responsible for constructing the RAG pipeline. It defines how the language model interacts with a retrieval mechanism to provide accurate and contextual responses. The script outlines the embedding generation, retrieval, and model configuration.

chatbot.py

This script serves as the entry point for interacting with the RAG chatbot. It defines the user interface, input/output handling, and logic for querying the RAG pipeline constructed in build_rag.py.

ollama_llm.py

This files provides and interface between OLLAMA and the langchain backend.

Dependencies

The project requires Python 3.8 or higher (On windows, can be installed via https://www.microsoft.com/store/productId/9NCVDN91XZQP?ocid=pdpshare), and Ollama (https://ollama.com/)

Recommend using llama3.2 (https://ollama.com/library/llama3.2) and/or codellama (https://ollama.com/library/codellama) models

For additional details w.r.t. LangChain, see (https://www.langchain.com/) for additional details, but the TLDR is that LangChain is a framework for developing applications powered by language models. It provides abstractions for chaining together multiple LLM calls and integrating them with external tools.

Usage

Python dependencies can be installed using the following command:

pip install -r requirements.txt

Build the RAG Pipeline

Run build_rag.py to set up the Retrieval-Augmented Generation pipeline. This step might include creating embeddings, configuring retrieval mechanisms, and ensuring the language model is properly integrated:

python3 build_rag.py --ouptut codebot.pkl ./

Run the Chatbot

Launch the chatbot interface by running chatbot.py:

python3 chatbot.py --vectorstore codebot.pkl

Notes

Scripts assume that you've already installed ollama and acquired a model (such as those recommended above), and that the ollama server is running (check via ollama ps to confirm).

See example_chat.xt for additional details (usage example)