Train Using Precomputed RPN Proposals?

[qinhuan]

Do you plan to support training Using Precomputed RPN Proposals?

[qinhaifangpku]

I have finished the code for training Using Precomputed RPN Proposals. Detectron: Fast & Mask R-CNN Baselines Using Precomputed RPN Proposals R-50-FPN 1x :33.7 Use this pytorch code: R-50-FPN 1x :33.76

And multi-scale train and test, and some code of PANet. My code is ugly, so I didn't organize the code and integrate it into this repo. https://github.com/qinhuan/mask-rcnn.pytorch_v0

Hi, do you know how to test the network with the given proposals?