SamsungLabs / fbrs_interactive_segmentation

[CVPR2020] f-BRS: Rethinking Backpropagating Refinement for Interactive Segmentation https://arxiv.org/abs/2001.10331

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Mixed precision training on hrnet32_ocr128.

aiwithshekhar opened this issue · comments

I tried training hrnet32_ocr128 with mixed precision, although the training time and memory on
gpu has reduced but there is no reduction in the inference speed.

  1. Can you suggest where I am going wrong.
  2. By what other meas I can decrease the inference speed.