Underdog Shows a 32MB Speech Model That Also Makes Music
Underdog has shown a new speech model that is only 32MB in size. Even though it is small, it can generate both speech and music. The demo shows that compact AI models are getting more capable, which matters for people who want AI tools that don't need large files.
Quick glossaryUnderdog, speech model
- Underdog
- Underdog is the company that built this small speech model; the story gives no further details about it.
- speech model
- An AI program trained to turn text into spoken voice or to understand spoken words.
Burning Potato score: 65%