search

KoljaB / LocalAIVoiceChat

Local AI talk with a custom voice based on Zephyr 7B model. Uses RealtimeSTT with faster_whisper for transcription and RealtimeTTS with Coqui XTTS for synthesis.
Other
392 stars 38 forks source link

Error with 'cloning_reference_wav' #6

Open ajeema opened 4 months ago

ajeema commented 4 months ago

All libraries seem to have been installed properly. However I get this error when trying to run

C:\Users\USER\code\LocalAIVoiceChat-main>start.bat cuda not available llama_cpp_lib: return llama_cpp Initializing LLM llama.cpp model ... llama.cpp model initialized Initializing TTS CoquiEngine ... Traceback (most recent call last): File "C:\Users\USER\code\LocalAIVoiceChat-main\ai_voicetalk_local.py", line 111, in <module> coqui_engine = CoquiEngine(cloning_reference_wav="female.wav", language="en", speed=1.0) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\Users\USER\AppData\Roaming\Python\Python311\site-packages\RealtimeTTS\engines\base_engine.py", line 11, in __call__ instance = super().__call__(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ TypeError: CoquiEngine.__init__() got an unexpected keyword argument 'cloning_reference_wav'

Please advise

ajeema commented 4 months ago

Removing it seems to have worked.

KoljaB commented 4 months ago

Probably update to latest RealtimeTTS version. Parameter cloning_reference_wav ist now named voice.