does this work without fine tuning

would you need to retrain the pretrained model to adapt to a particular speaker or can you test it on a input audio and it will do tts using the pretrained model?

[sdsantiagodiez]

Have you tried if model retraining is a must?