48K supports the whispering Shao Luoyin

Currently, the

Shaoluoyin model is trained using the Mygf-f048k base model, and a brand-new training method is employed to train this model.

I will share the advantages of this model

1: Support whispering.

2: Excellent emotional expression (but does not support shouting, etc.).

3: Accurate enunciation.

It's recommended to set your pitch around 17 or 16. If your original input sounds soft, then 11 or 12 is most suitable. The higher the pitch, the less ideal the emotional expression, but the closer it gets to the preview tone.

Whisper: This model supports whisper chatting, but it must be able to make sure the AI understands what it's saying. Here's a preview for everyone.

This model works even better when used with racks.