[D] Building a ML system vs project.

I personally spend my time split between traditional software engineer and machine learning.

Every time I am building or working on a model my OCD kicks in, notebooks, scripts, no real abstractions etc. All the ML code feels like it may be thrown away until it works and at that point I may as well keep the model created from the garbage code.

Having read and asked questions in the past I think many of us suffer with this, and …

abstractions building code engineer etc every machine machine learning machinelearning notebooks project scripts software software engineer spend