I’m running an object detection pipeline on a Hailo module with Raspberry Pi 5, and I’d like to monitor how much of the Hailo accelerator is actually being utilized. Specifically, I want to check:
The real-time utilization of the Hailo module when inference is running.
How many TOPS are being used by my object detection model.
Is there any recommended tool, SDK command, or API function that can report Hailo module utilization and TOPS usage during runtime?
Any guidance or example commands/scripts would be very helpful.
Before running inference open terminal and type the following command hailortcli monitor
And run your model in another terminal.
You can able to see utilization.