Prediction takes too long time on CPU

Community Help
1

Describe your question/issue here! (delete this when you post)
I am running inference on a local server on a CPU in windows. The server is running inside a podman container. for 400 images it takes ~6min for the prediction. Is there a way I can reduce the duration. Is it possible to check chunks of images?

  • Project Type:
  • Operating System & Browser:
  • Project Universe Link or Workspace/Project ID:
2

Hi @Anunaya_Thakur ,

Can you please share more information:

  • what is the model you are running
  • what is the size of images you are processing

Could you also try to run outside of container, directly in WSL, and confirm timings?

Thanks!
Grzegorz

3

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.