Skip to content
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.

Sign up for GitHub

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

Jump to bottom

[ENH] Reduce the number of backends #95

Open
pbarbarant opened this issue Oct 30, 2024 · 3 comments · May be fixed by #127
Open

[ENH] Reduce the number of backends #95

pbarbarant opened this issue Oct 30, 2024 · 3 comments · May be fixed by #127
Labels
enhancement New feature or request

Comments

@pbarbarant
Copy link
Collaborator

As for now, we rely on POT and Jax for classical OT and torch for fugw. In order to lighten up the package and ease the maintaining, we should cut out one or more backends.
@bthirion
Copy link
Contributor

bthirion commented Nov 3, 2024

Well, I'm still unsure which one we should keep...

@pbarbarant
Copy link
Collaborator Author

pbarbarant commented Feb 12, 2025 •
edited
Loading

I'd be in favor of removing the jax backend. The jit compilation of the sinkhorn kernels is actually slower than running the actual algorithm with POT. Since parcels have almost always different sizes, kernels can't be reused between them. Moreover now that SparseOT is implemented, ott's GPU support is not an advantage anymore.

@bthirion @emdupre What do you think?

@bthirion
Copy link
Contributor

I have no special interest is this backend, hence if this is better for you I'm fine with removing it.

@pbarbarant pbarbarant linked a pull request Feb 12, 2025 that will close this issue
Open
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
enhancement New feature or request
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

[BACKEND] Remove jax and make POT the default OT implementation Parietal-INRIA/fmralign
2 participants
@bthirion @pbarbarant