Open loboere opened 9 months ago
Maybe in terms of singing, but it's still a bit lacking in converting voice
Onthe RVC website, it says the training data is an English speaking dataset, as far as I've seen in these sovits 5 projects it's a singing dataset and I believe Korean, so there would be no surprise that RVC would be better for voice without a lot of extra training.
I would like to know if this work offers better results than RVC?