Open
Description
I have been looking at https://github.com/rhasspy/wyoming-whisper-cpp/ which is used for Home Assistant speech to text. It uses the deprecated main
binary to pipe audio samples into whisper.cpp.
It appears that whisper-cli
does not retain this functionality, and all the examples I've seen of eg piping ffmpeg into whisper.cpp use this main binary.
What would be the correct approach to implement a Wyoming server?
It does appear there is a https://github.com/ser/wyoming-whisper-api-client proxy server which seems a bit wasteful but might work.
Metadata
Assignees
Labels
No labels
Activity