Closed hdmjdp closed 1 year ago
Because it is a different model The ASR repo was made for StarGANv2-VC initially, while StyleTTS uses a different set of phonemes (IPA). I haven't updated the ASR repo yet, I will do so after I finish cleaning up the StyleTTS 2 code.
hi, https://github.com/yl4579/StyleTTS/blob/eac67158fd21578ea950ff4d6734bf6e6434b6b5/meldataset.py#L28C5-L28C5
the pretrained ASR model use this ""AA0",10 "AA1",11 "AA2",12 "AE0",13 "AE1",14 "AE2",15 "AH0",16 "AH1",17 "AH2",18 "AO0",19 "AO1",20 "AO2",21 "AW0",22 "AW1",23",
Why?