search

sh-lee-prml / HierSpeechpp

The official implementation of HierSpeech++
MIT License
1.17k stars 134 forks source link

Training code #6

Closed pawanhv closed 8 months ago

pawanhv commented 10 months ago

Hi, Will the training code be released to test it on our own datasets?

Pavan

sh-lee-prml commented 10 months ago

Thanks for your interest

We will have a plan to share the training code after paper acceptance or getting good reviews. We would appreciate your understanding.

Thanks Sang-Hoon

pawanhv commented 10 months ago

Ok thanks, good implementation thought 👍 do you have a timeline for the paper review ?

sh-lee-prml commented 10 months ago

Thanks for your understanding!

We submitted this paper to journal earlier this month so we don't know the exact timeline. If we receive the review, I will keep you posted on how it goes.

jupinter commented 9 months ago

How is everything going on? Looking forward to the release.

Pydataman commented 9 months ago

mark

mysxs commented 9 months ago

Hope to go online training code as soon as possible~

zhuangzhuangliu2345 commented 8 months ago

How is everything going on? Looking forward to the release.

youralmight commented 8 months ago

mark

hayeong0 commented 8 months ago

TTV training code

We have updated the Text-To-Vec training code. The training code for the speech synthesizer will be made public after the journal process of this research is completed. However, you can use a pre-trained our hierspeechpp synthesizer and train TTV with your own dataset. Our speech synthesizer is powerful in zero-shot, so after TTV training, it can synthesize high-quality speech. Thank you for taking an interest in our work! :)