Where do I host an optimizer?

Beginners
1

Hi, I have developed an optimization technique for adam called ladam (pip install ladam) that uses a laplacian and is useful when neighbors matter such as in some PINN, and I would like to know the best place to host this information on HF. I could not find an optimizer section anywhere.

TIA

Greg

1 Like
2

As for the Optimizer, I don’t think there’s any problem with putting it in the HF repository, but that just lets you store Python files and such, so it’s not very practical…
Also, if you want to integrate your optimizer into the Hugging Face ecosystem directly, the best approach is probably to commit it to the TRL GitHub repository.

Alternatively, the following method is commonly used:

HF doesn’t really have a dedicated “optimizer” section. The Hub’s first-class repo types are Models, Datasets, and Spaces, so for an optimizer library I’d usually keep the code/package on GitHub + PyPI and use HF for the parts that fit naturally.

Concretely, I’d suggest:

  • Space for an interactive demo or benchmark explorer
  • Show and Tell for the announcement / discussion thread
  • Dataset repo only if you want to publish benchmark traces or reproducibility data
  • Collection to group the Space + dataset + any paper in one place
  • Community Blog as a later long-form writeup, not the primary home

So if it were me: GitHub/PyPI as the canonical home, HF Space as the demo, then a Show and Tell post linking everything.

Community Blog could also work if you want a longer technical article, but I’d treat that as a follow-up: HF says PRO includes Community Blogs, and the current flow goes through Blog Explorers / new-blog.

1 Like
3

Thank you for these suggestions, will check them out now. Is there a general reason there is no optimizer section?

1 Like
4

Hmm… To put it simply, I guess it’s because “GitHub is already there for that purpose.”

On Hugging Face, Spaces is where you’d typically host code (for execution), but Spaces is mainly intended for demos.

Hugging Face Hub is like GitHub for files that are too large, such as AI model weights and massive datasets.

Presumably, Hugging Face assumes that users will use it alongside GitHub; they likely expect that code and libraries—including those developed by Hugging Face itself—will be hosted on GitHub.

1 Like
5

Thank you so much, this makes sense and is a great analogy.

1 Like
6

This topic was automatically closed 12 hours after the last reply. New replies are no longer allowed.