Distributed training for eager few-shot object detection

[Atia]

Is there a way to adapt this object detection approach for distributed training like we did in course 2? Any resources will be much appreciated.