MME (Multimodal Evaluation)

eval
BAAI
Comprehensive benchmark for vision-language models.
Paper (arXiv)GitHub
benchmarkmultimodalevaluation