Support coco val dataset with BLIP2 captions and VLRM captions

Question

Closed this issue a month ago · 0 comments