search

TensorSpeech / TensorFlowTTS

:stuck_out_tongue_closed_eyes: TensorFlowTTS: Real-Time State-of-the-art Speech Synthesis for Tensorflow 2 (supported including English, French, Korean, Chinese, German and Easy to adapt for other languages)
https://tensorspeech.github.io/TensorFlowTTS/
Apache License 2.0
3.84k stars 814 forks source link

How to use waveglow vocoder? #174

Closed LLianJJun closed 4 years ago

LLianJJun commented 4 years ago

Hi~

I want to check voice by combining your fastspeech, fastspeech2 and nvidia-waveglow waveglow

When training waveglow, if I train with min ~ max (80~7600), (default values 0 ~ 8000)
is it possible to combine with your txt2mel results?

dathudeptrai commented 4 years ago

@LLianJJun you should to check all parameter related with preprocessing steps. (hop_size, f0 min-max, ...). That mean you should check if the raw melspectrogram is the same for both model. My model can be revert to raw mels based-on mel stats pre-computed.

LLianJJun commented 4 years ago

Thnaks!

aragorntheking commented 3 years ago

@LLianJJun Did you succeed at this ? Can you share details ?