The idea is based on this article, the code and data come from this repo.
I want to start from this state of code and makes it "more MLOps", i.e.
- version control (Git and DVC),
- main code put outside notebook, to a proper Python package,
- testing (Pytest) and CI/CD (e.g. GitHub Actions),
- logging,
- configuration file for global variables (e.g. with dynaconf),
- Object-Oriented Programming with best practices, such as Design Patterns.
Install required dependencies:
pip install -r requirements.txt
Run the tests:
pytest tests.py
Run the main script:
python main_script.py