Simple RAG (Retrieval Augemented Generation)

This project is a lightweight web app for question-answering over PDFs using Retrieval Augmented Generation (RAG). Users upload a PDF, ask questions, and get answers based on the document content via a Hugging Face model.

Table of Contents

• Prerequirements
• Project Structure
• Model
• Architecture
• Features
• Installation
• How to get API Access Token
• Usage
• License

Prerequirements

• or above: Download here
• : Sign up here

Project Structure

simple-RAG/
├── static/
│   ├── style.css
│   └── script.js
├── templates/
│   └── index.html
├── .gitignore
├── server.py
├── worker.py
├── .env.example
├── Dockerfile
├── requirements.txt
├── LICENSE
└── README.md

Model

• HuggingFaceTB/SmolLM3-3B (You can use different models)

Architecture

flowchart TD
    A[User #40;Web Browser#41;] -- Prompt --> B[Flask Web Server<br/>#40;server.py#41;]
    B --> C[Worker<br/>#40;worker.py#41;]
    F[PDF file] -- Store --> D
    C --> D[Vector DB<br/>#40;Chroma#41;]
    C --> E[LLM]
    A -- Upload --> F
    D -- Retrieves context --> C
    E -- Generates answer --> C
    C -- Responds --> B
    B -- Returns answer --> A

Loading

Features

• Upload and query PDF documents
• RAG-powered chatbot interface
• Uses Hugging Face models
• Local semantic search with Chroma
• Light/dark mode UI

Installation

To install this project, open your Terminal and follow these steps:

1. Clone the repository:

   $ git clone https://github.com/arthurtran04/simple-RAG.git

2. Change the directory to simple-RAG:

   $ cd "$(find . -type d -name "simple-RAG")"

3. Create a Python virtual environment .venv and install the required dependencies:

   $ python -m venv .venv
   source .venv/bin/activate
   pip install -r requirements.txt

4. Set up environment variables:

   $ cp .env.example .env

5. Configure your .env file:

   # Hugging Face API Token
   HF_TOKEN=

How to get API Access Token

1. Go to Hugging Face website, click your avatar and click "Access Tokens"

2. Click "Create new token"

3. Name your token and tick the two checkboxes in the "Inference" section

4. Scroll down and click "Create token"

5. Copy your Access Token before closing and paste it into the HF_TOKEN variable inside the .env file

Usage

To start the application, run the server.py file:

$ python server.py

This application will run locally at http://127.0.0.1:5000:

Upload your PDF file and enter your prompt in the textbox below, and the chatbot will respond:

To stop the application, use Ctrl + C in the Terminal

License

This project is licensed under the Apache-2.0 License. See the LICENSE file for more details.