Textless NLP: missing tacotron2 training script for GSLM

Question

Textless NLP: missing tacotron2 training script for GSLM

Opened this issue 3 years ago · 8 comments

🐛 Bug

There is no script provided to train the tacatron2 in the following path: fairseq/examples/textless_nlp/gslm/unit2speech/tacotron2

Can we please get the training script as well so that we are able to train our own decoder (modified tacotron2) on a new dataset?

Thanks!

Answer 1 · 2021-10-21T14:57:33.000Z

@hikushalhere Can you please help us here?

Answer 2 · 2021-10-31T07:56:22.000Z

@eugene-kharitonov could you please help with this? Thank you!

Answer 3 · 2021-11-09T17:22:10.000Z

Can someone please help with this? Thanks

Answer 4 · 2021-11-10T06:50:53.000Z

As far as I remember, the training code for modified Tacotron2 was not open sourced with GSLM. @wnhsu, can you help?

Answer 5 · 2021-11-16T10:06:51.000Z

As far as I remember, the training code for modified Tacotron2 was not open sourced with GSLM. @wnhsu, can you help?

Thank you @hikushalhere for the update. Looking forward to @wnhsu's input on this.

Answer 6 · 2021-11-17T18:32:27.000Z

@ahazeemi you can refer to the fairseq unit-to-speech example here, or the NVIDIA's Tacotron2 repo link.

Answer 7 · 2021-11-17T20:35:55.000Z

@ahazeemi you can refer to the fairseq unit-to-speech example here, or the NVIDIA's Tacotron2 repo link.

Thank you for the reply. After training the Transformer/FastSpeech2 model on units (as described here), can that be used in place of the modified tacotron in the GSLM speech resynthesis?

Answer 8 · 2021-11-17T23:15:05.000Z

I don't think you can replace that since it's a reimplementation