#Finally got my full voice assistant pipeline working
1 messages · Page 1 of 1 (latest)
quite a nice pipeline
linux-voice-assistant (webcam mic -> wake word) -> speaches (whisper speech to text) -> voice assisstant ( home assistant logic, calls qwen LLM via ollama if it cant understand) -> speaches again (kokoro tts) -> linux-voice-assistant again (laptop speakers)
all in a shitload of docker containers
The last comment made me laugh so thanks for that