Configure chunk size in streaming mic capture


Is there a way to configure the chunk size in mic capture with streaming=True?

I am using gr.Audio(source=‘microphone’) for capturing audio from mic and then perform real time speech to text. Model supports 160ms as smallest chunk size. But I see that the handler function gets called with 1 second audio chunks and there is no way to control chunk size. This impacts the latency in showing real time transcripts.