Your question is Repository Structure for Reproducibility. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
What are the best practices for structuring a repository to ensure modularity and reproducibility for MSD machine learning work?