Videostf Github Io

A normal output describes the video coherently. A repetitive output collapses into loops of the same phrases or sentences. Video Large Language Models (VideoLLMs) are usually judged by what they predict.

Description read from the page itself, because the listing it was found in gave none.