MMMU-Benchmark / MMMU

This repo contains evaluation code for the paper "MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGI"
https://mmmu-benchmark.github.io/
Apache License 2.0
353 stars 24 forks source link

Link for the open source methods in Leaderboard #14

Closed XiongweiWu closed 6 months ago

XiongweiWu commented 9 months ago

Hi I find that the leaderboard has added some new methods, however, for some open source methods, I cannot find either their codes or even their papers. Can u add the relevant link to these methods (e.g., Macro-VL, InfiMM-zephyr) on your leaderboard?

xiangyue9607 commented 9 months ago

Thanks! The authors haven't released their models yet but they said they would release them very soon. We will add the link shortly.