NVIDIA / RULER

This repo contains the source code for RULER: What’s the Real Context Size of Your Long-Context Language Models?
Apache License 2.0
728 stars 47 forks source link

Request for permissions #61

Closed ChenAlmagor closed 1 month ago

ChenAlmagor commented 2 months ago

Hi @hsiehjackson, Could you please grant me the necessary access to open PRs? I would like to contribute AI21 Jamba model results to the leaderboard and add it as one of the supported models. Thanks!

hsiehjackson commented 2 months ago

Hi @ChenAlmagor I am not sure what permission you need. You are welcome to open any PRs! BTW, I update our main table to show AI21 Jamba-1.5-mini results. It is a pretty strong model!!

yuvalbelfer commented 2 months ago

Hey @hsiehjackson , Thanks for adding Jamba-1.5-Mini! Can you also add Jamba-1.5-Large? The results are also pretty strong (see section 6.3.1 in our whitepaper).

hsiehjackson commented 2 months ago

Sure! Sorry I don't have enough compute to run on my side, so I directly copy numbers from your report!

yuvalbelfer commented 2 months ago

Thank you!