How I Split My ML App into Microservices (Because a Monolithic Dockerfile Was a Bad Idea) (opens in new tab)

When developing Machine Learning applications, the initial focus is almost always on training the model and looking at validation scores…