Chat PDF

Chatbot for chatting with PDFs via a RAG system that provides the LLM with only the relevant document chunks. A chat memory is built in, allowing long discussions with the document.

---

---

---

Structure

src/
└── rag
    ├── __init__.py     # Global variables of app
    ├── __main__.py     # Entry point of app
    ├── cli.py          # Entry point of Typer CLI app
    ├── app.py          # Entry point of FastAPI app
    ├── routes.py       # Routes of FastAPI app
    ├── pipeline.py     # Langchain chain functions
    ├── interface.py    # Chainlit interface definition

Installation

You must have installed python3.12 with pip. This project use uv as a virtual environment manager. To install it, run the following command:

pip install --upgrade pip
pip install uv

Then, create a virtual environment and install the dependencies:

uv sync

If your IDE don't create the virtual environment automatically, you can create it manually with the following command:

uv venv
source .venv/bin/activate

After this you must copy paste the .env.example file to .env and fill it with your own values.

cp .env.example .env

Usage

You can run the CLI with the following command:

uv run src/rag

python src/rag

Tests

You can run the tests with the following command:

pytest

You will find coverage report in the htmlcov folder, just open the index.html file in your browser.

Linting

This project use ruff for formatting and linting. You can launch both with uv custom command. To do this, run the following command:

uv run lint

Production

This project use Docker, docker-compose and docker tools for publish and deploy the app.

You can run in local the container with the following command:

docker compose -f compose.prod.yml up --build

You can update the preprod or prod with the following command:

make update-preprod