Evaluation benchmarks used to evaluate vision-language models

Hello - What evaluation benchmarks used to evaluate image-description matching in vision-language models?

Have you heard of METEOR ?

Hi I have but not for vLM.