voicegen / text /symbols.py
Plachta's picture
Upload 66 files
c6070db
raw
history blame
2.27 kB
'''
Defines the set of symbols used in text input to the model.
'''
# japanese_cleaners
_pad = '_'
_punctuation = ',.!?-'
_letters = 'AEINOQUabdefghijkmnoprstuvwyzʃʧ↓↑ '
'''# japanese_cleaners2
_pad = '_'
_punctuation = ',.!?-~…'
_letters = 'AEINOQUabdefghijkmnoprstuvwyzʃʧʦ↓↑ '
'''
'''# korean_cleaners
_pad = '_'
_punctuation = ',.!?…~'
_letters = 'ㄱㄴㄷㄹㅁㅂㅅㅇㅈㅊㅋㅌㅍㅎㄲㄸㅃㅆㅉㅏㅓㅗㅜㅡㅣㅐㅔ '
'''
'''# chinese_cleaners
_pad = '_'
_punctuation = ',。!?—…'
_letters = 'ㄅㄆㄇㄈㄉㄊㄋㄌㄍㄎㄏㄐㄑㄒㄓㄔㄕㄖㄗㄘㄙㄚㄛㄜㄝㄞㄟㄠㄡㄢㄣㄤㄥㄦㄧㄨㄩˉˊˇˋ˙ '
'''
'''# zh_ja_mixture_cleaners
_pad = '_'
_punctuation = ',.!?-~…'
_letters = 'AEINOQUabdefghijklmnoprstuvwyzʃʧʦɯɹəɥ⁼ʰ`→↓↑ '
'''
'''# sanskrit_cleaners
_pad = '_'
_punctuation = '।'
_letters = 'ँंःअआइईउऊऋएऐओऔकखगघङचछजझञटठडढणतथदधनपफबभमयरलळवशषसहऽािीुूृॄेैोौ्ॠॢ '
'''
'''# cjks_cleaners
_pad = '_'
_punctuation = ',.!?-~…'
_letters = 'NQabdefghijklmnopstuvwxyzʃʧʥʦɯɹəɥçɸɾβŋɦː⁼ʰ`^#*=→↓↑ '
'''
'''# thai_cleaners
_pad = '_'
_punctuation = '.!? '
_letters = 'กขฃคฆงจฉชซฌญฎฏฐฑฒณดตถทธนบปผฝพฟภมยรฤลวศษสหฬอฮฯะัาำิีึืุูเแโใไๅๆ็่้๊๋์'
'''
'''# cjke_cleaners2
_pad = '_'
_punctuation = ',.!?-~…'
_letters = 'NQabdefghijklmnopstuvwxyzɑæʃʑçɯɪɔɛɹðəɫɥɸʊɾʒθβŋɦ⁼ʰ`^#*=ˈˌ→↓↑ '
'''
'''# shanghainese_cleaners
_pad = '_'
_punctuation = ',.!?…'
_letters = 'abdfghiklmnopstuvyzøŋȵɑɔɕəɤɦɪɿʑʔʰ̩̃ᴀᴇ15678 '
'''
'''# chinese_dialect_cleaners
_pad = '_'
_punctuation = ',.!?~…─'
_letters = '#Nabdefghijklmnoprstuvwxyzæçøŋœȵɐɑɒɓɔɕɗɘəɚɛɜɣɤɦɪɭɯɵɷɸɻɾɿʂʅʊʋʌʏʑʔʦʮʰʷˀː˥˦˧˨˩̥̩̃̚ᴀᴇ↑↓∅ⱼ '
'''
# Export all symbols:
symbols = [_pad] + list(_punctuation) + list(_letters)
# Special symbol ids
SPACE_ID = symbols.index(" ")