Leaderboard runtime error and request for pending model evaluations

#9
by adirik - opened

Hi @malhajar thank you for the great work on the leaderboard!

The leaderboard seems to be down with a runtime error at the moment. I wanted to kindly ask if you could restart the app and also inquire when you might have time to evaluate the pending models? A few of them (including mine - neuralwork/gemma-2-9b-it-tr) are finetuned versions of the highest-ranking models, so it'd be great to see how they compare.

My own evaluation yielded competitive results, but I wasn't able to benchmark using vLLM, so I’m refraining from posting self-reported metrics.

Hi @adirik , Thanks for reaching out. I have been in a small break due to new year lately so i couldn't attend to the leadboard.
I have upgraded the leadboard ( Using my own resources :D ) i beilieve we won't recieve any more runtime errors.
I will do the evaluations and post them today.

Thanks!

Thank you, greatly appreciate the effort you put into maintaining the leaderboard :)

adirik changed discussion status to closed

Sign up or log in to comment