Can I use jiant for multi input task learning like (image, audio, video)?

[Imran-tech-lab]

Hi, I want to use jiant library for multi task learning, like gpt4, Blip2, model.
If yes I can do this, so please give me sample notebook example.