tensorRT workspace size definition #6472
-
Hey, I was going through the export.py code and was wondering what exactly is Workspace argument since it says "Some tactics do not have sufficient workspace memory to run. Increasing workspace size may increase performance, please check verbose output" when I run the code to convert .pt weights to .engine. Thanks |
Beta Was this translation helpful? Give feedback.
Answered by
glenn-jocher
Jan 30, 2022
Replies: 1 comment 3 replies
-
@ausollet default workspace size is 4GB. You can increase this with the python export.py --include tensorrt --workspace 8 |
Beta Was this translation helpful? Give feedback.
3 replies
Answer selected by
ausollet
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
@ausollet default workspace size is 4GB. You can increase this with the
--workspace
argument, i.e. for 8GB: