myshell-ai / MeloTTS

High-quality multi-lingual text-to-speech library by MyShell.ai. Support English, Spanish, French, Chinese, Japanese and Korean.
MIT License
4.84k stars 631 forks source link

What is the sampling rate of the pre-trained model? #191

Closed gansidui closed 1 month ago

gansidui commented 2 months ago

The training document recommends a sampling rate of 44100, but I provide data with a sampling rate of 24000 for training. Will there be any difference in the effect? I want to keep the sampling rate consistent with the pre-trained model instead of resampling it.

justinatbahasa commented 1 month ago

hi @gansidui how's the result training with 24khz audio?

gansidui commented 1 month ago

你好@gansidui使用 24khz 音频进行训练的效果如何?

The effect is pretty good. During training, 24Khz will be resampled and the final result is still 44100Hz.