Speed up transformation #1462
Unanswered
rubick1896
asked this question in
Q&A
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
I found the transform which happens on the CPU takes more time than the actual prediction on GPU for a trained DeeapAR network. In the inference phase, the prediction time is bottlenecked by the transform. What are some possible ways to speed up the transform? Maybe allowing multiple cores to do the job?
Beta Was this translation helpful? Give feedback.
All reactions