MaxMax2016

https://github.com/mozilla/TTS/tree/master/TTS/speaker_encoder

python app.py 就是启动webUI吧

是的，这是专门用于转换歌声的

FreeVC

https://github.com/PlayVoice/Grad-SVC/blob/20230920-V2-96/hubert/hubert_model.py ``` def consume_prefix(state_dict, prefix: str) -> None: keys = sorted(state_dict.keys()) for key in keys: if key.startswith(prefix): newkey = key[len(prefix):] state_dict[newkey] = state_dict.pop(key) def hubert_soft( path: str, ) -> HubertSoft:...

do you have any dump logs?

maybe you can try RMVPE: https://github.com/thestmitsuki/so-vits-svc-rmvpe https://github.com/DLSeed/so-vits-svc-5.0

do you use whisper large v2?

不能合并，我试过，合并出来的音质很差。

是的