FastSpeech2
FastSpeech2 copied to clipboard
For aishell3 dataset, the MFA g2p dictonary is not incompatible with Fastspeech2/pinyin-lexicon-r.txt.
For aishell3 dataset, the MFA g2p dictonary is not incompatible with Fastspeech2/pinyin-lexicon-r.txt.
Could you provide a solution for this bug? such as providing the aishell3 acoustic model.
run “mfa train” instead of “mfa align”