AI_REPO

This repository contains code for chatting with a GitHub Repo that leverages an Open Source large language model (LLM) to provide informative responses to user queries.

Project Architecture

Demo

Assignment_Demo.mp4

Prerequisites

Before running the code, please ensure you have the following installed:

• Python 3.7 or later (3.10 was used in the project)

API tokens for:

• Replicate (LLM provider)
• Activeloop
• GitHub

Setup

Clone this repository:

git clone https://github.com/<your-username>/AI_Repo.git

Navigate to the repository directory:

cd <repo-name>

Create a .env file:

In the root of the repository, create a file named .env.
Add the following lines to the .env file, replacing the placeholders with your actual API tokens:

REPLICATE_API_TOKEN=<your_replicate_token>
ACTIVELOOP_TOKEN=<your_activeloop_token>
GITHUB_TOKEN=<your_github_token>
DATASET_PATH=<path_to_store_vector_store>

Running the Code

Install required libraries:

pip install -r requirements.txt

Launch the Gradio interface:

python main.py

Interact with the system

• A Gradio interface will open in your web browser.
• Type your question in the chat box and press Enter.
• The system will process your query and provide a response using the LLM.

Additional Notes

The code is currently configured to use the segment-anything GitHub repository as the data source. If you want to use a different repository, modify the github_url variable in the main function.

You can adjust the LLM model, vector store settings, and other parameters in the code as needed.