andabi / deep-voice-conversion

Deep neural networks for voice conversion (voice style transfer) in Tensorflow
MIT License
3.92k stars 845 forks source link

other dataset try,eg ljspeech #118

Open CaoRong2016 opened 5 years ago

CaoRong2016 commented 5 years ago

hello ~ I wonder if default params is suitable for ljspeech dataset ? could anyone give some tips for other dataset training. the result is really different between cmu_slt and other dataset with default params