Add video-text-to-text task category, MIT license, and link VG-GUI-Bench paper

Add README.md

Upload dataset

initial commit