Docs
Browse documentation

Docs / Speech

Speech models

Every released Phonon repository, its sizes, and its measured word error rates.

Phonon-1

The default speech recognition model and the one behind fermion transcribe, listen, and speech serving. It transcribes English from 16 kHz mono audio and was trained at 2.4 bits per weight from the start. The aliases phonon, stt, and asr all resolve to it. Benchmark figures are word error rate; lower is better.

DownloadOn diskLibriSpeech test-cleanLibriSpeech test-otherEight-benchmark macro
415 MB455 MB2.645.707.67
Open Phonon-1 on Hugging Face

Phonon-1 Micro

The smallest speech model. It trades accuracy for a smaller download and is selected with --model micro.

DownloadOn diskLibriSpeech test-cleanLibriSpeech test-otherEight-benchmark macro
285 MB331 MB3.006.518.52
Open Phonon-1-Micro on Hugging Face

Phonon-1-Big

DownloadOn diskLibriSpeech test-cleanLibriSpeech test-otherEight-benchmark macro
581 MB822 MB2.675.727.60
Open Phonon-1-Big on Hugging Face