Caching ML models, this might take a while 😅