caizexin / tf_multispeakerTTS_fc

the Tensorflow version of multi-speaker TTS training with feedback constraint

caizexin/tf_multispeakerTTS_fc Issues

Question about the speaker embedding visualization by t-SNE?
Updated 2 years ago1
why not preprocess wav to fbank before train speaker embedding? it is too long that training while processing
Updated 2 years ago1
Question regarding training hardware details
Updated 4 years ago
Problem running the speaker embedding network while training from scratch.
Updated 4 years ago8
__init__() missing 2 required positional arguments: 'fbanks' and 'labels'
Updated 4 years ago2