OpenVoice
OpenVoice copied to clipboard
Fix audio model initialization in se_extractor.py with cpu support
I think this needs to be implemented urgently