Best Dataset for Ai Vocals?

As time goes on, things get improved

So far I heard about RVC, so-vits-svc or diff-svc, any of these any good for Ai Singing/Rapping?

Im not sure, which one to pick.

I’m open to other suggestions.