Diarization in stream example doesn't seem to work

[dimon777]

First off, huge thanks and kudos to all devs who made this amazing tool! Thank you!
Now, I am testing simple diarization in stream example:

./build/bin/whisper-stream -m models/ggml-small.en-tdrz.bin -t 4  -tdrz

And it doesn't seem to show//recognize different speakers. The output text just goes as a whole w/o any speaker markers. Anything I do wrong?