Docker project with Streamlit, FastAPI and SQLite

This project is part of a course in Text Mining, an exciting discipline that explores text analysis techniques for extracting meaningful information. Carried out in a school context, this project aims to apply the concepts and methodologies learned in class to a concrete application.

Prerequisites

Make sure you have Docker and Docker Compose installed on your machine.

• Docker
• Docker Compose

Clone project

git clone https://github.com/moiseberthe/sise-nlp.git

Project structure

• client/: Application Streamlit
  • controller/
  • pages/
  • app.py
  • Dockerfile
  • env.sample.py
  • requirements.txt
• server/: Application FastAPI
  • database/
  • utils/
  • app.py
  • Dockerfile
  • env.sample.py
  • migrate.py
  • requirements.txt

Environment configuration

Before running the application, you need to create two env.py files, one in the server/ directory and another in the client/ directory. You can take inspiration from the available example files.

For the client :

cp client/env.sample.py client/env.py 

For the server :

cp server/env.sample.py server/env.py 

You need to have credentials to access the Pôle Emploi API, which is used in this project to retrieve job offers. Visit [https://pole-emploi.io](pole-emploi.io) to obtain these credentials.

This is optional, but you will only have access to job offers available on Apec.

Docker setup

The Docker files (client/Dockerfile and server/Dockerfile) contain the build configurations for the Streamlit and FastAPI services.

The docker-compose.yml file is designed to orchestrate the different services.

Build and run containers

To build and start the services using Docker Compose, you can use the following command:

docker compose up --build

The project will be accessible at the following address:

• Streamlit: http://localhost:8501
• FastAPI: http://localhost:8000

Push images on Docker Hub

Server

docker tag nlp-server:latest moiseberthe/nlp-server:latest

docker push moiseberthe/nlp-server:latest

Client

docker tag nlp-client:latest moiseberthe/nlp-client:latest

docker push moiseberthe/nlp-client:latest