mayfool

> > Hi @GuangChen2016 , thanks for your attention. Yes, I met the same issue and I'm still figuring out what's going on. In my opinion, it may be from...

> > > > Hi @GuangChen2016 , thanks for your attention. Yes, I met the same issue and I'm still figuring out what's going on. In my opinion, it may...

> Hi @mayfool, can you show the mel-spectrogram image which has a single frequency line ? All synthesised wavs have single frequency line. Not few of them. So I think...

> > > Hi @mayfool, can you show the mel-spectrogram image which has a single frequency line ? > > > > > > All synthesised wavs have single frequency...

Thanks for you reply, I will have a try!

I think the current pretrained model doesn't support zero-shot voice-cloning? We need to train from the start with the speak embedding replaced with embedding vector extracted from spkrec-ecapa-voxceleb? https://github.com/collabora/WhisperSpeech/blob/934a67c74c8d23d233f9702d51c5be149411a304/whisperspeech/s2a_delar_mup_wds.py#L567C1-L567C56