mirror of
https://github.com/tcsenpai/agenticSeek.git
synced 2025-06-06 11:05:26 +00:00
42d2655881
FIX:
llm_provider.py 102: _model -> model
MOD:
utility.py 25: add windows termial color print
text_to_speech.py 29: add windows version
ADD:
add termcolor in requirements.txt
44 lines
1.6 KiB
Python
44 lines
1.6 KiB
Python
from kokoro import KPipeline
|
|
from IPython.display import display, Audio
|
|
import soundfile as sf
|
|
import subprocess
|
|
import re
|
|
import platform
|
|
|
|
|
|
class Speech():
|
|
def __init__(self, language = "english") -> None:
|
|
self.lang_map = {
|
|
"english": 'a', # 🇺🇸 'a' => American English
|
|
"chinese": 'z', # 🇯🇵 'j' => Japanese: pip install misaki[ja]
|
|
"japanese": 'j' # # 🇨🇳 'z' => Mandarin Chinese: pip install misaki[zh]
|
|
}
|
|
self.pipeline = KPipeline(lang_code=self.lang_map[language])
|
|
|
|
def speak(self, sentence):
|
|
sentence = self.clean_sentence(sentence)
|
|
generator = self.pipeline(
|
|
sentence, voice='af_heart', # <= change voice here
|
|
speed=1, split_pattern=r'\n+'
|
|
)
|
|
for i, (gs, ps, audio) in enumerate(generator):
|
|
audio_file = 'sample.wav'
|
|
print(audio_file)
|
|
display(Audio(data=audio, rate=24000, autoplay=i==0))
|
|
sf.write(audio_file, audio, 24000) # save each audio file
|
|
if platform.system().lower() != "windows":
|
|
subprocess.call(["afplay", audio_file])
|
|
else:
|
|
import winsound
|
|
winsound.PlaySound(audio_file, winsound.SND_FILENAME)
|
|
|
|
def clean_sentence(self, sentence):
|
|
sentence = re.sub(r'`.*?`', '', sentence)
|
|
sentence = re.sub(r'[^a-zA-Z0-9.,!? ]+', '', sentence)
|
|
sentence = re.sub(r'\s+', ' ', sentence).strip()
|
|
return sentence
|
|
|
|
if __name__ == "__main__":
|
|
speech = Speech()
|
|
speech.speak("hello would you like coffee ?")
|