Instrumental/Singing Voice only Generation

[shoegazerstella]

Hi all, this is a great project and I'm having lots of fun experimenting with it!

My question is: Would it be possible to force the model to only generate instrumental parts or singing voice only?
I am trying the continuation from an instrumental song of mine, without any lyrics, and I notice sometimes there's some random singing.
Do you think this could be done? Maybe by fine-tuning on voice-only tracks, do you think the generation, in this case, would only be random singing without any instrumental background?
Thanks!

[svntv]

Great feature! I was wondering if there is a parameter that can be set or maybe putting an empty lyrics or something like that?

[LucarioGamer]

I would love to see this but as you know OpenAI makes a cool ai but abandons it and leaves it super unoptimized.

[Cortexelus]

You could use OpenMix or Spleeter or Wave-U-Net to remove the vocals (works on generative music too)