RAG Demo for Sun* TechUp 2024

Introduction

This is a demo for the Retrieval Augmented Generation (RAG) model. The RAG model is a combination of a retriever and a generator. The retriever is responsible for finding relevant documents from a large corpus, and the generator is responsible for generating the answer based on the retrieved documents.

Installation

To install the required libraries, run the following command:

pip install -r requirements.txt

Usage

Config the OpenAI Azure

You need to config the Azure OpenAI API key in the azure_config.py file.

Indexing the documents

First you need push the documents into the /data folder. Then you can index the documents by running the following command:

python populate_database.py

After this step the database will be save into /chroma folder

Running the RAG model

After indexing the documents, you can run the RAG model by running the following command:

python query_data.py "Why did Mr. Kobayashi Taihei come to Vietnam?"

Run with GUI interface

python ui.py