🎭 EmoAct-MiMo: Emotion-Controllable Text-to-Speech

Generate intensely emotional speech using the EmoAct-MiMo model.

This is still a very early experiment and is very early in the training run, I need to change a few settings and retrain. But the model turned out quite nicely!

It may hallucinate, try a few times to get good results.

Voice cloning is not supported yet.
Model Selection

v1.0 is the current stable model. v1.1 is a beta experimental version with potentially different characteristics.
Examples
Model Selection Emotion Text