Whisper speech-to-text service
A daemon that provides speech-to-text transcription using OpenAI's Whisper model.
Supports GPU acceleration with NVIDIA CUDA for faster inference.
Configure the service with:
snap set whisper-stt port=9000
snap set whisper-stt model=base
Available models: tiny, base, small, medium, large
For GPU support, connect the GPU interface:
snap connect whisper-stt:gpu-2404 mesa-2404:gpu-2404