Is fusion_encoder getting used for video captioning?

[Astuary]

Hello!

I was looking at model_video_caption_mplug.py and saw that self.fusion_encoder is not used in the forward pass.

Do we need to instantiate it for the video captioning tasks?