2 ms·

Thanks - I was wondering more about the translation pipeline - i am assuming something like whisper and m2m100? How did you manage to get the latency so low? Ot

Thanks - I was wondering more about the translation pipeline - i am assuming something like whisper and m2m100? How did you manage to get the latency so low? Other examples I have seen feel really sluggish in comparison.

billylo 2y ago
Fairly basic there... chunking audio (websocket) and send them for translation sequentially.