04N: Sequence Reconstruction using FMs and RAG

Project Overview

This project aims to develop a pipeline for the reconstruction of sequential information such as chemical synthesis steps or timeline information using Foundation Models (FMs) and Retrieval Augmented Generation (RAG). The project focuses on enabling users to understand the flow of sequential information and quickly identify any gaps in the sequence, specifically targeting domains like chemical synthesis.

External Industry Client/Advisor: NVIDIA

Project Structure

Deco-3801/
│
├── myenv/
├── src/
│   └── agent/
│       ├── .env
│       ├── graph.py
│       ├── frontend.py
|       ├── app.py
│       ├── ingest.py
│       ├── langgraph.json
│       └── requirements.txt
│       ├── styles.css
├── web_scrapers/
└── README.md

Installation

1. Set up Virtual Environment

First, create and activate a virtual environment in the root directory:

cd Deco-3801
python3 -m venv venv
source venv/bin/activate (MacOS)

./venv/Scripts/activate (Windows)

2. Install Project Dependencies

Navigate to the src/agent directory and install the required dependencies:

cd src/agent
pip install -r requirements.txt

How to run the Project

1. Create a .env file

Create a .env file inside agent

2. Add environment variables inside the .env file

NVIDIA_API_KEY=
TAVILY_API_KEY=

create nvidia nim api key by clicking here.

create tavily api key by clicking here.

3. Start the backend server

open a new terminal

cd src/agent (MacOS)
python app.py

python -m src.agent.app (Windows)

4. Start the frontend

open a new terminal

cd src/agent (MacOS)
streamlit run frontend.py

streamlit run .\src\agent\frontend.py (Windows)

LangGraph Studio Instructions

While in Beta, LangGraph Studio is available for free to all LangSmith users on any plan tier. # for LangSmith here.

1. Install Docker Desktop

Ensure that Docker Desktop is installed on your machine. You can download and install it from the official website.

2. Install LangGraph Studio

Download the latest .dmg file of LangGraph Studio by clicking here. Currently, only macOS is supported. Also depends on Docker Engine to be running.

3. Start LangGraph Studio

Once LangGraph Studio is installed, start it in the Deco-3801/src/agent directory.