Why the motiondecoder is able to decode both text and motion?

[wang-zm18]

https://github.com/athn-nik/teach/blob/f4285aff0fd556a5b46518a751fc90825d91e68b/teach/model/teach.py#L295

I think it is necessary to differetiate the text decoder and the motion decoder. So I am confused. Thank you in advance!

[athn-nik]

Can you please elaborate on this? The motiondecoder is responsible for decoding the texts and the motion latents.