metadata

title: RAG Implementation with Ollama and LangChain
colorFrom: red
colorTo: blue
emoji: 🤗
sdk_version: 4.37.1
sdk: gradio
app_file: my_ollama.py
pinned: false

RAG Implementation with Ollama and LangChain

This repository contains an implementation of the Retrieval-Augmented Generation (RAG) model using Open AI's LLaMA large language model and the LangChain library.

Overview

The code instantiates an OLLAMA model, loads Wikipedia content, splits the text into manageable chunks, creates sentence embeddings with SentenceTransformers, and builds a vector store using Chroma. Finally, it creates a QA chain using the OLLAMA model and the vector store retriever.

Usage

To use this code, simply run the Python script. The output will be the generated response to a given question.

Example Question

The example question used in this implementation is: "What is Tsai's energy policy?"

Dependencies

LangChain

OLLAMA

SentenceTransformers

Chroma

Notes

This is my first attempt at implementing RAG using OLLAMA and LangChain. While the code is functional, it may not be optimized for performance or scalability. Further improvements and testing are needed to ensure the model's reliability.

I hope this helps! Let me know if you have any questions or need further assistance.