An implementation of Microsoft's "FastSpeech 2: Fast and High-Quality End-to-End Text to Speech"
Geek Repo:Geek Repo
Github PK Tool:Github PK Tool
malu01 opened this issue a year ago · comments
I noticed that the sampling frequency of LibriTTS data set is 24000, but 22050 was set in your configuration file. Did you downsample the data before the training? Or did you just use the raw data?