Precise RoI Pooling with coordinate gradient support, proposed in the paper "Acquisition of Localization Confidence for Accurate Object Detection" (https://arxiv.org/abs/1807.11590).
So under pytorch1.0, it seems once you run code and it will re-compile PrPooling module, is there any way to use compiled module and doesn't need re-compile anymore in future run?
So under pytorch1.0, it seems once you run code and it will re-compile PrPooling module, is there any way to use compiled module and doesn't need re-compile anymore in future run?