forked from killua/TakwayDisplayPlatform
13 lines
777 B
Python
13 lines
777 B
Python
|
from utils.bert_vits2_utils import TextToSpeech
|
|||
|
|
import soundfile as sf
|
|||
|
|
tts = TextToSpeech()
|
|||
|
|
tts.print_speakers_info()
|
|||
|
|
|
|||
|
|
audio, sample_rate= tts.synthesize("你好，我好开心", # 文本
|
|||
|
|
0, # 说话人 id
|
|||
|
|
style_text="我很难过！！！！呜呜呜！！！", # 情绪prompt，当language=="ZH" 才有效
|
|||
|
|
style_weight=0.9, # 情绪prompt权重
|
|||
|
|
language="mix", # 语言类型，包括 "ZH" "EN" "mix"
|
|||
|
|
en_ratio=1.) # mix语言类型下，英文文本速度，越大速度越慢
|
|||
|
|
save_path = "./tmp2.wav"
|
|||
|
|
sf.write(save_path, audio, sample_rate)
|