Training for encoder (i.e. embedding) usage only?

auspicious3000 / SpeechSplit

Unsupervised Speech Decomposition Via Triple Information Bottleneck

http://arxiv.org/abs/2004.11284

MIT License

636 stars 92 forks source link

Training for encoder (i.e. embedding) usage only? #60

Closed tobwei closed 2 years ago

tobwei commented 2 years ago

Sorry I just found your e-mail address and will thus just write you an e-mail about this - this can be deleted - sorry again for the inconveniences.

jinny1208 commented 2 years ago

hi, just wondering what your question was since I am also interested in just using the encoder part only and not the decoder. And, if you don't mind, can you also post the reply to your question? Thanks in advance.