Chat PDF

Presentation

This project is example Chat-PDF, a web application that lets you chat with any PDF. This project uses LangChain for the prompt engineering, Gradio for the web user interface and Gradio to create the CLI for production.

Installation

This project use poetry, to install the dependencies, you can run the following command:

poetry install

Usage

To run the project, you can use the following command:

poetry run python src

Structure

├── src               # Project source code
├── docs              # Project documentation
│   └── static        # README.md static files
├── tests             # Folder containing software tests
│   ├── units         # Unit tests
│   └── integrations  # Integration tests
├── scripts           # Useful scripts for the project (no CI/CD)
├── ruff.toml         # Ruff configuration file
├── environment.yml   # Conda environment configuration file

Installation

This project requires conda to be installed. To install the dependencies, simply run the following command:

conda env create -f environment.yml

You can update the environment with the following command:

conda env update -f environment.yml

Usage

This project uses typer to create a command-line interface. To launch command help, simply issue the following command the following command:

python src

Docker

Docker-compose

This project have a docker-compose.yml file to run the project in production. To run the project, you can use the following command:

• build: Build the Docker image if not exists
• d: Mode daemon (run in background)

docker-compose up --build -d

If the build Docker timeout because miniconda install is too long, you can use the following command:

COMPOSE_HTTP_TIMEOUT=3600 sudo docker compose up --build

Manual

First, you must create the volume who allow Airflow to access to Gradio's 'db' folder contains the PDF content in vector

docker volume create shared_db_volume

To build the Docker image, you can run the following command:

docker build -t chat_pdf .

Gradio app

To build the Dockerfile, you can use the following command:

docker build -t gradio_chat_pdf .

To run the Docker container, you can use the following command:

• d: Mode daemon
• p: Port to use
• v: set volume for Airflow container (clear the db folder)
• name: give name to container and don't have random name

docker run -d --name gradio_chat_pdf -v shared_db_volume:/app/db -p 7860:7860 gradio_chat_pdf

Airflow app

To build the Dockerfile, go to airflow folder and use the following command:

docker build -t airflow_chat_pdf .

To run the Docker container, you can use the following command:

• d: Mode daemon
• p: Port to use
• v: set volume for Airflow container (clear the db folder)
• name: give name to container and don't have random name

docker run -d --name airflow_chat_pdf -v shared_db_volume:/opt/airflow/db -p 8080:8080 airflow_chat_pdf

Tips & Other

For debugging purposes, you can run the following command:

docker run -it -p 7860:7860 xxxx_chat_pdf /bin/bash

For deleting all Docker containers, you can use the following command:

docker rm -f $(docker ps -a -q)

For deleting all Docker images, you can use the following command:

docker rmi -f $(docker images -q)

Airflow

This project uses Apache Airflow to schedule the data processing. To start the Airflow web server, you can run the following command:

airflow db init
airflow db migrate

Create an admin user:

airflow users create \
    --username admin \
    --firstname FIRST_NAME \
    --lastname LAST_NAME \
    --role Admin \
    --email [email protected] \
    --password admin

To start the Airflow web server in development, you can run the following command:

airflow standalone

To start the Airflow web server in production, you can run the following command:

airflow webserver --port 8080 --daemon
airflow scheduler --daemon