search

yl4579 / AuxiliaryASR

Joint CTC-S2S Phoneme-level ASR for Voice Conversion and TTS (Text-Mel Alignment)
MIT License
111 stars 30 forks source link

how to make word_index_dict.txt #3

Closed Ruinmou closed 2 years ago

Ruinmou commented 2 years ago

I have a little immature question, how to make word_index_dict.txt about Mandarin?

Ruinmou commented 2 years ago

My understanding is to write all the appearing phonemes into the txt file and correspond to an id, is my understanding correct?

Charlottecuc commented 2 years ago

I guess the first five tokens should be "pad", "sos", "eos", "unk" and blank. e.g. https://github.com/yl4579/AuxiliaryASR/blob/5cb3ee46c4839d96d50c8ad9b3f378b1d7f06df7/models.py#L106

Ruinmou commented 2 years ago

Thanks, got it