Spaces:

fair-forward
/

languagebench

Running

Mismatch between languages and benchmark datasets

by OrianeN - opened 7 days ago

I like to use your leaderboard from time to time to get an idea of the best systems for certain languages, esp. French and Occitan.

However, it seems like the Occitan leaderboard is based on datasets for which there is no data in this language... For instance, your app indicates that it uses CohereLabs/Global-MMLU, although this language is not covered by the dataset.

Could you have a look and maybe fix this ?

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment