AIが動く。未来が更新される。Phonon-2
Phonon-2 is the most accurate open speech recognition model for English under 900 MB. Across the Open ASR Leaderboard's seven English sets it averages 5.21 % word error, beating models multiple times its size in raw bytes. Set for set it holds the accuracy of the 2.5 GB full-precision parakeet-tdt-0.6b-v3 it derives from, reaching 100.8 % of that model's word accuracy on parliamentary speech and beating it on meetings, from a download 15 times smaller. It transcribes an hour of audio in about 20 seconds on an M5 MacBook Air (174x realtime), at 143x on eight Zen 5 cores (16 vCPU) and at 6,680x on one H100 in batches of 128. † Open ASR Leaderboard's published row; the other rows use its code on the full test sets. * Size from the parameter count at 16 bits.
翻訳準備中のため原文を表示しています
Phonon-2 is the most accurate open speech recognition model for English under 900 MB. Across the Open ASR Leaderboard's seven English sets it averages 5.21 % word error, beating models multiple times its size in raw bytes. Set for set it holds the accuracy of the 2.5 GB full-precision parakeet-tdt-0.6b-v3 it derives from, reaching 100.8 % of that model's word accuracy on parliamentary speech and beating it on meetings, from a download 15 times smaller. It transcribes an hour of audio in about 20 seconds on an M5 MacBook Air (174x realtime), at 143x on eight Zen 5 cores (16 vCPU) and at 6,680x on one H100 in batches of 128. † Open ASR Leaderboard's published row; the other rows use its code on the full test sets. * Size from the parameter count at 16 bits.