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

CQADupstackProgrammersRetrieval results are missing for a lot of models #63

Open
x-tabdeveloping opened this issue Dec 6, 2024 · 2 comments

Comments

@x-tabdeveloping
Copy link
Contributor

A lot of models in the MTEB(eng, classic) tab are missing CQADupstackProgrammersRetrieval results and therefore get nan mean, and don't get displayed in the plot.
I checked and the files are missing here, in the results repo. Is this something we should run (@Muennighoff) or something we can fetch from the old results (@Samoed )?

@Samoed
Copy link
Collaborator

Samoed commented Dec 6, 2024

I think this should be runned, because curently when creating model card, then all CQADupstack datasets results are combined
https://github.com/embeddings-benchmark/mteb/blob/a6ce6f9b7050c1fad60e0c6e8985afa9356e2728/mteb/create_meta.py#L104

@KennethEnevoldsen
Copy link
Contributor

Yep. I actually also think this will leave to a difference in the mean (which is a bit annoying). A solution is to allow custom aggregators for some benchmarks (might also fix the issue with displaying bright)

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

No branches or pull requests

3 participants