f197177ae1
- FastAPI 백엔드: /api/voices, /api/tts, /api/health - 한국어 메인 + 영어(5종 악센트), 중국어/일본어 UI 제거 - 언어별 목소리 모델 선택, 속도(0.5~2.0x)/피치(±12반음) 조절 - 로그인/사용제한/과금 유도 없음 (무제한 테스트) - NVIDIA GPU 자동 사용 (torch cu128, Blackwell sm_120) - 유저 친화 한국어 UI (frontend) - Docker/compose (GPU 예약), 8788 포트 - MeloTTS(순수 파이썬) 벤더링 + 검증된 버전 고정(constraints)
36 lines
1.4 KiB
Python
36 lines
1.4 KiB
Python
from .symbols import *
|
|
|
|
|
|
_symbol_to_id = {s: i for i, s in enumerate(symbols)}
|
|
|
|
|
|
def cleaned_text_to_sequence(cleaned_text, tones, language, symbol_to_id=None):
|
|
"""Converts a string of text to a sequence of IDs corresponding to the symbols in the text.
|
|
Args:
|
|
text: string to convert to a sequence
|
|
Returns:
|
|
List of integers corresponding to the symbols in the text
|
|
"""
|
|
symbol_to_id_map = symbol_to_id if symbol_to_id else _symbol_to_id
|
|
phones = [symbol_to_id_map[symbol] for symbol in cleaned_text]
|
|
tone_start = language_tone_start_map[language]
|
|
tones = [i + tone_start for i in tones]
|
|
lang_id = language_id_map[language]
|
|
lang_ids = [lang_id for i in phones]
|
|
return phones, tones, lang_ids
|
|
|
|
|
|
def get_bert(norm_text, word2ph, language, device):
|
|
from .chinese_bert import get_bert_feature as zh_bert
|
|
from .english_bert import get_bert_feature as en_bert
|
|
from .japanese_bert import get_bert_feature as jp_bert
|
|
from .chinese_mix import get_bert_feature as zh_mix_en_bert
|
|
from .spanish_bert import get_bert_feature as sp_bert
|
|
from .french_bert import get_bert_feature as fr_bert
|
|
from .korean import get_bert_feature as kr_bert
|
|
|
|
lang_bert_func_map = {"ZH": zh_bert, "EN": en_bert, "JP": jp_bert, 'ZH_MIX_EN': zh_mix_en_bert,
|
|
'FR': fr_bert, 'SP': sp_bert, 'ES': sp_bert, "KR": kr_bert}
|
|
bert = lang_bert_func_map[language](norm_text, word2ph, device)
|
|
return bert
|