bacow/GPT-SoVITS
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

more code refactor

Browse Source
This commit is contained in:
Blaise
2024-01-16 17:14:18 +01:00
parent 0d92575115
commit 0d3d47f3c3
44 changed files with 4516 additions and 2623 deletions
Download Patch File Download Diff File Expand all files Collapse all files

30
GPT_SoVITS/AR/text_processing/phonemizer.py
View File

@@ -27,46 +27,44 @@ class GruutPhonemizer:
"—": "—",
"…": "… ",
"«": "«",
"»": "»"
"»": "»",
}
self._punctuation_regexp: str = rf"([{''.join(self._special_cases_dict.keys())}])"
self._punctuation_regexp: str = (
rf"([{''.join(self._special_cases_dict.keys())}])"
)
def _normalize_punctuation(self, text: str) -> str:
text = regex.sub(fr"\pZ+{self._punctuation_regexp}", r"\1", text)
text = regex.sub(fr"{self._punctuation_regexp}(\pL)", r"\1 \2", text)
text = regex.sub(rf"\pZ+{self._punctuation_regexp}", r"\1", text)
text = regex.sub(rf"{self._punctuation_regexp}(\pL)", r"\1 \2", text)
text = regex.sub(r"\pZ+", r" ", text)
return text.strip()
def _convert_punctuation(self, word: Word) -> str:
if not word.phonemes:
return ''
if word.phonemes[0] in ['‖', '|']:
return ""
if word.phonemes[0] in ["‖", "|"]:
return word.text.strip()
phonemes = ''.join(word.phonemes)
phonemes = "".join(word.phonemes)
# remove modifier characters ˈˌː with regex
phonemes = re.sub(r'[ˈˌː͡]', '', phonemes)
phonemes = re.sub(r"[ˈˌː͡]", "", phonemes)
return phonemes.strip()
def phonemize(self, text: str, espeak: bool=False) -> str:
def phonemize(self, text: str, espeak: bool = False) -> str:
text_to_phonemize: str = self._normalize_punctuation(text)
sents: List[Sentence] = [
sent
for sent in self._phonemizer(
text_to_phonemize, lang="en-us", espeak=espeak)
for sent in self._phonemizer(text_to_phonemize, lang="en-us", espeak=espeak)
]
words: List[str] = [
self._convert_punctuation(word) for word in itertools.chain(*sents)
]
return ' '.join(words)
return " ".join(words)
def transform(self, phonemes):
# convert phonemes to ids
# dictionary is in symbols.py
return [
self.symbol_to_id[p] for p in phonemes
if p in self.symbol_to_id.keys()
]
return [self.symbol_to_id[p] for p in phonemes if p in self.symbol_to_id.keys()]
if __name__ == "__main__":