search

microsoft / SpeechT5

Unified-Modal Speech-Text Pre-Training for Spoken Language Processing
MIT License
1.16k stars 113 forks source link

No code for Speech Synthesis #8

Closed petervickers closed 2 years ago

petervickers commented 2 years ago

Code for finetuning speech synthesis with the predicted log Mel-filterbank features, as described in the SpeechT5 paper, is not availiable.

Is it possible to provide this?

Many thanks

Ajyy commented 2 years ago

Hi, @mechanicalsea is preparing the code of TTS. Please stay tuned for further updates! Thanks!

mechanicalsea commented 2 years ago

Hi, @mechanicalsea is preparing the code of TTS. Please stay tuned for further updates! Thanks!

Got it. I will prepare the code of TTS and release it as soon as possible.

mechanicalsea commented 2 years ago

Code for finetuning speech synthesis with the predicted log Mel-filterbank features, as described in the SpeechT5 paper, is not availiable.

Is it possible to provide this?

Many thanks

Hello, the TTS code and scripts are prepared after this commit.

The scripts and manifest can be helpful to implement SpeechT5 TTS. And we also did implement it, and it can generate frequent speech utterances.

Thanks for you attention to our work.

mechanicalsea commented 2 years ago

Hi, @petervickers We provide additional re-implementation of TTS fine-tuned model speecht5_tts.pt and generated samples, but with a smaller batch size or max updates, which can be helpful.