How to compress ML models, clearly explained (with code): ...