Running the models in python/cli

[sina-mansour]

Hi Considering the relatively fast and efficient implementation of the models, I was interested to try and run these as part of some pipeline.

First of all, do you think that's recommended, or would you suggest other tools.

If this is recommended, is there a way to expose the pretrained models to a standalone python/bash script? Essentially, giving a T1 image as input and saving the output segmentation resulting from the deep NN?