Skip to content

Sequential Inference (Non-MP) #2705

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Jun 2, 2025

Conversation

LennyN95
Copy link
Contributor

This PR resolves #2556.
Improved implementation following #2614.

  • Implementation of preprocessing based on a single process
  • Introduce new environment variables nnUNet_npp and nnUNet_nps to set -npp and -nps
  • Using the new use predict_from_files_sequential function

We have done some internal testing and get identical results between the latest nnunetv2==2.5.1 and with our proposed patch.

@LennyN95
Copy link
Contributor Author

LennyN95 commented Mar 4, 2025

@FabianIsensee, did you find a moment to look into this? The changes are minimal but necessary to use your new sequential function from the CLI. Let me know if there's anything I can do to expedite the process or assist in any other way.

Once resolved, we can proceed with implementing the following models:

@LennyN95
Copy link
Contributor Author

@FabianIsensee Hi Fabian, I know times are busy. This PR includes minimal changes to enable the use of your new non-parallel inference from the CLI. I would appreciate it if you could take a quick look or suggest a rough timeline.

Thank you and best, Leo.

@FabianIsensee
Copy link
Member

Hey, sorry for being so slow 😅. Looks great, thank you for this addition! Many will be happy to have this!

@FabianIsensee FabianIsensee merged commit e561842 into MIC-DKFZ:master Jun 2, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Optional Torch Multiprocessing in nnUNet for Improved Security and Compatibility
2 participants