Unable to recreate MMLU-Pro Score

#34

by acostea-ionos - opened 5 days ago

5 days ago

Hello, we where unable to recreate the MMLU-Pro score. Our results gives us a score of 47% (using TIGER-LAB dataset) while the score in the model card is 66.3%. Are you able to tell us how the test was done? Or if there is a typo of any sort going on?

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

Tap or paste here to upload images

Your need to confirm your account before you can post a new comment.

· Sign up or log in to comment