Spaces:

autoevaluate

/

leaderboards

like 137

Running

Feedback from a community member @mrm8488

I do not know if you already do it but "bigger" models will tend to show better results so maybe in the leaderboard i'd show the model size

WDYT @Tristan - does this make sense to you to include? I know there's criticism of using parameter count as a proxy for "better", but maybe it's an interesting piece of knowledge to know when picking a model?