about a confused code comment "When it detects end of speech (non-voice for 500ms)"

[lq0104]

when I see the code whisper_online.py, I watch a line comment:

whisper_streaming/whisper_online.py

Line 523 in 225f038

When it detects end of speech (non-voice for 500ms), it makes OnlineASRProcessor to end the utterance immediately.

My question is: I cannot see any variable about the value of 500 in the MAIN branch. I just see it in the "vad-steaming" branch:

whisper_streaming/voice_activity_controller.py

Line 8 in fdd3091

min_silence_to_final_ms = 500,

. So is there something missing in the code merge, or is there a misunderstanding on my part?

[Gldkslfmsd]

Thanks. It looks like you found a bug. I thought it was 500ms but it seems to be 100ms: https://github.com/ufal/whisper_streaming/blob/225f0383553a23f4ecc4c4751b73bc406a120c6c/silero_vad.py#L13C18-L13C41

So let's make it 500ms here:

whisper_streaming/whisper_online.py

Line 538 in 225f038

self.vac = VADIterator(model) # we use all the default options: 500ms silence, etc.

and test whether it's doing what we expect.