LLM Apps using Streamlit and LangChain

Introduction

This repository showcases a suite of experimental LLM applications built using Streamlit.

Uses

• Streamlit (UI)
• LangChain
• OpenAI API (gpt-3.5-turbo)

Run the App

Clone the Repository

$ git clone github.com/rexsimiloluwah/streamlit-llm-apps
$ cd streamlit-llm-apps

Install the dependencies

You can advisably create a virtual environment

$ pip install -r requirements.txt

Run the app

$ streamlit run src/main.py

# Using make
$ make run-app

Example Applications

1. Simple Document QA App

This application enables you to perform question-answering over your PDF document. It uses the RetrievalQA chain and the in-memory DocArray vector store provided by LangChain.

2. Web Page QA App

This application enables you to perform question-answering over content loaded from a web page. It similarly uses the RetrievalQA chain and the in-memory DocArray vector store provided by LangChain.

3. Document Chat App

This application enables you to chat over your PDF document. It uses the ConversationalRetrievalChain chain and the in-memory DocArray vector store provided by LangChain. The memory is managed externally.