tiny-voca
tiny-voca Brings Browser-Based Speech Synthesis with Just 4.5M Parameters
tiny-voca is a fully open-source, 4.5M-parameter English TTS model that runs locally in browsers at about 70x real time.
2 articles tagged with this topic
tiny-voca is a fully open-source, 4.5M-parameter English TTS model that runs locally in browsers at about 70x real time.
nari-labs' open-source TTS hits 50ms response (faster than a blink). For short-video creators: near-zero wait, time saved, money saved.