Great thanks to the reference project of this work: apeatling/ollama-voice-mac
- Install Ollama.
- Download the a model, Mistral 7b is a actually a ood choice using the
ollama pull mistral
command. - Download an OpenAI Whisper Model and place the
.pt
model file in a/whisper
directory in the repo root folder. - Make sure you have FFmpeg by running
brew install ffmpeg
. - For Apple silicon support of the PyAudio library you'll need to install Homebrew and run
brew install portaudio
. - Run
poetry install
to install all dependencies (pip install -r requirements.txt
also works). - Run
python chat.py
to start the assistant.