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.

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

🚀[FEA]: Add Kolmogorov - Arnold Networks for Modulus #494

Open
androbomb opened this issue May 13, 2024 · 5 comments
Open

🚀[FEA]: Add Kolmogorov - Arnold Networks for Modulus #494

androbomb opened this issue May 13, 2024 · 5 comments
Assignees
Labels
? - Needs Triage Need team to review and classify enhancement New feature or request

Comments

@androbomb
Copy link

Is this a new feature, an improvement, or a change to existing functionality?

New Feature

How would you describe the priority of this feature request

Low (would be nice)

Please provide a clear description of problem you would like to solve.

Due to its usefulness in Scientific Machine Learning, it would be relevant to add Kolmogorov - Arnold Network Layers in Modulus.

Describe any alternatives you have considered

No response

@androbomb androbomb added ? - Needs Triage Need team to review and classify enhancement New feature or request labels May 13, 2024
@sohale
Copy link

sohale commented Jun 4, 2024

I am interested to do this. Is there a process to plan for it? Or should one just go and come back with a PR that assumes the API design (choice of arguments, etc) they like?

PS. I see there is one available already: https://github.com/androbomb/kan-modulus

@mnabian
Copy link
Collaborator

mnabian commented Oct 18, 2024

Hi @androbomb, would you be interested in contributing your code here to Modulus? I'll be happy to work with you to get your PR merged.

@mnabian mnabian self-assigned this Oct 18, 2024
@androbomb
Copy link
Author

Hi @androbomb, would you be interested in contributing your code here to Modulus? I'll be happy to work with you to get your PR merged.

Dear @mnabian , I would love to. Let me know how to properly do it. Furthermore, I have started working on FastKAN implementation [see ArXiV paper], where instead of B-Splies, Jacobi/Chebyshev polinomials, a superposition of (learnable) Gaussian Kernels are used.

Nevertheless, I must stress that the pure PyTorch implementations I used in my github are adaptation of others users code, as reported in the code and in the Readme.md:

Let me know how to proceed.
Have a nice day

@androbomb
Copy link
Author

androbomb commented Oct 21, 2024

Hi @androbomb, would you be interested in contributing your code here to Modulus? I'll be happy to work with you to get your PR merged.

Dear @mnabian , I updated my GitHub folder adding the code for FastKAN (KAN with Radial Basis Function instead of B-splines).

In that repository I have created Modulus SYM models (not plain Modulus). Should I create a PR for Modulus SYM?
If so, can I add you to the fork as Maintainer to help me creating the PR?

Best regards

@mnabian
Copy link
Collaborator

mnabian commented Jan 14, 2025

Hi @androbomb, sorry I missed this thread. Yes please go ahead and open a PR. Feel free to add me as maintainer to your fork.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
? - Needs Triage Need team to review and classify enhancement New feature or request
Projects
None yet
Development

No branches or pull requests

3 participants