Adding device choice for whisper model to constructor

KoljaB / RealtimeSTT

A robust, efficient, low-latency speech-to-text library with advanced voice activity detection, wake word activation and instant transcription.

MIT License

2.1k stars 191 forks source link

Adding device choice for whisper model to constructor #60

Closed vancoder1 closed 5 months ago

vancoder1 commented 5 months ago

Since I had vram memory overflowing issue on CUDA while running it alongside with RealtimeTTS on GPU I decided to add the device choice to constructor for people with low-end PCs

KoljaB commented 5 months ago

Looks good,, thanks a lot