fix: Installation tests by pinning lm-eval extras.

[ParagEkbote]

Description

Installation tests have been failing on main due to missing wheels for nvidia-cutlass-dsl-libs-base and nvidia-cudnn-frontend since both of these deps publish wheels only for linux, not for MacOS and Windows. Could you please review?

cc: @johannaSommer

ref:
https://github.com/PrunaAI/pruna/actions/runs/23354770495/job/67942686570
https://github.com/PrunaAI/pruna/actions/runs/23354770495/job/67942686541

Related Issue

Fixes #(issue number)

Type of Change

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

How Has This Been Tested?

Checklist

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes

Additional Notes

[ParagEkbote]

I'm a bit perplexed about the deps either being extras of vllm or lm-eval. Could you please let me know?

cc: @johannaSommer

[sdiazlor]

Hi @ParagEkbote! Thanks for the quick review. We'll check asap! It should solve the failings.

[davidberenstein1957]

Closed and re-opened to re-trigger the tests :)

[github-actions]

This PR has been inactive for 10 days and is now marked as stale.

[begumcig]

I am a bit confused, why did we change the vllm extras, for fixing the lm-eval problem?

[ParagEkbote]

Hi, as you can view in the failing jobs, there are a couple of missing deps which are not installed during the installation test, which may cause the test to fail. Attaching some images for more context:

Image 1:

Image 2:

So, I am a bit uncertain if the deps should be a part of vllm or lm-eval dependency grouping since these are cuda driver packages which are generally pulled for vllm. I'd be happy to iterate to change them for lm-eval dependency grouping.