hello everyone!

i have been building Refrakt for the past few months, a workflow for training and evaluating computer vision models.

deep learning models today are fragmented: