Generative Spoken Language Model based on continuous word-sized audio tokens

Connections are not prepared for this record yet, or its source metadata has changed. The original record remains available while background snapshots are rebuilt.