|
|
@@ -81,7 +81,7 @@ import torchaudio
|
|
|
cosyvoice = CosyVoice('speech_tts/CosyVoice-300M-SFT')
|
|
|
# sft usage
|
|
|
print(cosyvoice.list_avaliable_spks())
|
|
|
-output = cosyvoice.inference_sft('你好，我是通义语音合成大模型，请问有什么可以帮您的吗？', '中文女')
|
|
|
+output = cosyvoice.inference_sft('你好，我是通义生成式语音大模型，请问有什么可以帮您的吗？', '中文女')
|
|
|
torchaudio.save('sft.wav', output['tts_speech'], 22050)
|
|
|
|
|
|
cosyvoice = CosyVoice('speech_tts/CosyVoice-300M')
|
lyuxiang.lx