Zama works on a specific product vision and executes deliverables on a quarterly basis. We have restricted ourselves to numpy and torch for the moment.
So, at this time, per se, you can’t natively compile a keras model in FHE but you can however turn the forward function into numpy, and the rest should follow successfully.
But if you are willing to contribute, please help! Please have a look here. If you can turn keras to numpy, this should work!