Suggestions

• Conventional commits: see https://www.conventionalcommits.org/en/v1.0.0/
• Branch development: use git branches to organize your work. Each task is a branch with a PR.

My version of public learning is creating this notebook of difficulties and what was done to overcome it. Hopefully, it can be useful to improve the course as well

Module 1

1 - Local Env Setup (macOs)

1.1. Check Local MacOS envprep by Adetbekov for Docker setup

1.2. Package Version Control using Poetry

• Install poetry: pip install poetry
• Start environemnt: poetry init which will create a pyproject.toml file with main dependencies. Feel free to skip dependency and install them later using poetry add
• Start a .venv poetry shell and add .venv to .gitignore

(Bonus) Updating code that uses seaborn.distplot

Module 2

Adding MLFlow

• Poetry already showed to be useful since if we use pip to install MLFlow it will install 2.1.3 by default which is incompatible with pyarrow 16
• Also, when using requirements.txt in the videos, make sure to define versions of packages. Example: my first version of sklearn was not supported with MLFlow's autologging

First test of running train.py with MLFlow autlogging

• 

• 

•