Machine Learning

Mini Project Structure

A clean ML project layout you can reuse.

Suggested Layout

project/
  data/                 # raw + processed
  notebooks/            # exploration only
  src/
    features.py         # preprocessing/feature logic
    train.py            # training entrypoint
    predict.py          # inference entrypoint
  models/               # saved artifacts
  tests/                # unit tests
  README.md

Rule of thumb

Keep notebooks for exploration; keep src/ for reusable code.
Save the full pipeline (not just the model weights).