Moshi is a speech-text foundation model and full-duplex spoken dialogue framework. It uses Mimi, a state-of-the-art streaming neural audio codec.

Yet Another Picoprobe

Some simple plugins for generating stator coils in kicad