This is a demo of real time speech to text with OpenAI's Whisper model. It works by constantly recording audio in a thread and concatenating the raw bytes over multiple recordings.
The project has been updated to use poetry.
PyAudio package has a dependency on PortAudio library, So, we need to
install the PortAudio library on your Mac. You can do this using the Homebrew
package manager with the following command:
brew install portaudio
Then, to install dependencies simply run
poetry install
Reference: see original installation info