search

NVIDIA / mellotron

Mellotron: a multispeaker voice synthesis model based on Tacotron 2 GST that can make a voice emote and sing without emotive or singing training data
BSD 3-Clause "New" or "Revised" License
855 stars 183 forks source link

Inference bug? #65

Closed lqniunjunlper closed 4 years ago

lqniunjunlper commented 4 years ago

While librosa.core.load() returns float array, these is no need to norm by MAX_WAV_VALUE?

rafaelvalle commented 4 years ago

Very nice catch, thank you! We've updated the repo.