Deploy your own model on Hailo

Hello everyone, my name is Mr. Jean-Yves. I apologize for the inconvenience. When I try to use Hailort to load my model into HEF, it displays errors such as [Hailort] [error] CHECK failed - Driver version (4.20.0) is different from library version (4.21.0)
[Hailort] [error] Driver version mismatch, status HAILO_INVALID_DRIVER_VERSION(76)
[Hailort] [error] CHECK_SUCCESS failed with status=HAILO_INVALID_DRIVER_VERSION(76) . When I did some research and searched your platform, I saw that you can use DeGirum. I would like to ask if you can load your own model with this module?

Hi @Jean_Yves_Kouadio1
The error you are seeing is due to mismatch between driver version and runtime version.
Regarding DeGirum: yes you can use your own model with DeGirum PySDK. Please let us know if you encounter any issues.

@Jean_Yves_Kouadio1 You must make sure the HailoRT, HailoPCIe driver, and PyHailoRT are of the same version. There is a concise step-by-step instruction how to get it working.

You don’t have to run the multi-modal multi-stream inference. For single chip, the installation procedure is identical.