TODOs for Effective ML teamwork at an early-stage startup - Machine Learns

Abstracting ML code sacrifices expressiveness, increases coupling, and aggravates maintenance. These might be ok for regular software. But things are different for ML. I am sure you know how it feels to waste hours trying to match the API when you want to implement an ML trick. APIs and abstractions are bad for fast-paced ML R&D. ML is too fast, and any API is outdated from its inception. We see a similar pattern with well-known ML libraries (Transition from Theano - Tensorflow - PyTorch - JAX…).