Closed hdmjdp closed 1 year ago
If we do not use pretrained ASR ckpt, just let the asr model update parameters according to TTS model. Can we get good TTS result?
Yes, but it would take more time, and you will need different learning rates for the TTS model and ASR model. See the hyperparameters of pre-training for the ASR model.
If we do not use pretrained ASR ckpt, just let the asr model update parameters according to TTS model. Can we get good TTS result?