tensorrtx icon indicating copy to clipboard operation
tensorrtx copied to clipboard

Published 20 hours ago • verified publisher icon wang-xinyu

multibatch in Retinaface

Open sungggat opened this issue 2 years ago • 2 comments

Env

  • GPU T4.
  • OS Ubuntu18.04.
  • Cuda 10.2
  • TensorRT version 7.0

About this repo

  • which branch/tag/commit are you using?
  • which model? retinaface

Your problem

  • what is your command? e.g. python retinaface_trt.py

I am trying to use multi-batch. I created engine with batchsize 8. But output is correct for the first image, for others output is zero. What parts of decode.cu and decode.h I have to change in order to use multi batch

sungggat avatar Apr 07 '22 13:04 sungggat

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

stale[bot] avatar Jun 06 '22 14:06 stale[bot]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

stale[bot] avatar Aug 05 '22 19:08 stale[bot]