Yhn SportBaseball first seen 10 d ago, last 10 d ago, peak #10
Phonon-2 open speech recognition model released at 164 MB
Original: Phonon-2: most accurate open speech recognition model in a 164 MB download
Fermion Research has released Phonon-2, an open speech recognition model it describes as the most accurate available in a 164 MB download. The compact size makes it notable for running on-device, and the release is drawing attention from developers comparing its accuracy against larger speech-to-text systems.
Why now: Developers are interested in a highly accurate, lightweight open speech recognition model that can run locally.
Evidence
- Phonon-2: most accurate open speech recognition model in a 164 MB download · handfuloflight · 7
API: https://socialmediatrends-api.osmike.com/v1/trends/652081