Provided model doesn't work [HAILO8]

[almex-m-bulbul]

We are trying to use the ' facial landmark detection model` on Hailo8 device, apparently it's not working.

model : tddfa_mobilenet_v1.hef
link: https://hailo-model-zoo.s3.eu-west-2.amazonaws.com/ModelZoo/Compiled/v2.10.0/hailo8/tddfa_mobilenet_v1.hef

log:

$ hailortcli benchmark .\tddfa_mobilenet_v1.hef
Starting Measurements...
Measuring FPS in hw_only mode
[HailoRT] [error] CHECK_AS_EXPECTED failed - Error, Could not find logical periph bytes per buffer value
[HailoRT] [error] CHECK_EXPECTED failed with status=HAILO_INVALID_ARGUMENT(2) - 
[HailoRT] [error] CHECK_EXPECTED_AS_STATUS failed with status=HAILO_INVALID_ARGUMENT(2) -
[HailoRT] [error] CHECK_SUCCESS failed with status=HAILO_INVALID_ARGUMENT(2) -  
[HailoRT] [error] CHECK_SUCCESS_AS_EXPECTED failed with status=HAILO_INVALID_ARGUMENT(2) -
[HailoRT] [error] CHECK_EXPECTED failed with status=HAILO_INVALID_ARGUMENT(2) - 
[HailoRT] [error] CHECK_EXPECTED failed with status=HAILO_INVALID_ARGUMENT(2) - 
[HailoRT] [error] CHECK_EXPECTED failed with status=HAILO_INVALID_ARGUMENT(2) - 
[HailoRT] [error] CHECK_EXPECTED failed with status=HAILO_INVALID_ARGUMENT(2) - 
[HailoRT] [error] CHECK_EXPECTED failed with status=HAILO_INVALID_ARGUMENT(2) - 
[HailoRT] [error] CHECK_EXPECTED failed with status=HAILO_INVALID_ARGUMENT(2) - 
[HailoRT CLI] [error] CHECK_EXPECTED failed with status=HAILO_INVALID_ARGUMENT(2) - Failed configure vdevice from hef
[HailoRT CLI] [error] CHECK_EXPECTED_AS_STATUS failed with status=HAILO_INVALID_ARGUMENT(2) - hw_only measuring failed

I have tried to compile the onnx model, the compilation goes smoothly, but while ruuning it shows the same error.

[omerwer]

Hi @almex-m-bulbul,
I took the HEF you provided and ran the command you did on my machine and it ran successfully.
What HailoRT are you using? do you have a Hailo chip connected?

Regards,

[almex-m-bulbul]

@omerwer Hailo chip is connected. Same issue has been addressed my other teammate on a different machine as well.

$ hailortcli scan  
Hailo Devices:
[-] Device: 0000:04:00.0

$ hailortcli --version               
HailoRT-CLI version 4.15.0

It's perfectly working when checked with other models

$ hailortcli benchmark repvgg_a0_person_reid_2048.hef
Starting Measurements...
Measuring FPS in hw_only mode
Network repvgg_a0_person_reid_2048/repvgg_a0_person_reid_2048: 100% | 15382 | FPS: 1025.34 | ETA: 00:00:00
Measuring FPS and Power in streaming mode
[HailoRT] [warning] Using the overcurrent protection dvm for power measurement will disable the overcurrent protection.
If only taking one measurement, the protection will resume automatically.
If doing continuous measurement, to enable overcurrent protection again you have to stop the power measurement on this dvm.
Network repvgg_a0_person_reid_2048/repvgg_a0_person_reid_2048: 100% | 15386 | FPS: 1025.34 | ETA: 00:00:00
Measuring HW Latency
Network repvgg_a0_person_reid_2048/repvgg_a0_person_reid_2048: 100% | 6335 | HW Latency: 1.93 ms | ETA: 00:00:00

=======
Summary
=======
FPS     (hw_only)                 = 1025.42
        (streaming)               = 1025.36
Latency (hw)                      = 1.93251 ms
Device 0000:04:00.0:
  Power in streaming mode (average) = 1.68351 W
                          (max)     = 1.70233 W

[omerwer]

Hi @almex-m-bulbul,
I checked it on a machine with HailoRT v4.16.0, so it might have been a bug that was fixed in the recent release.
Can you please try to upgrade the driver + lib file (sudo dpkg -i hailort-pcie-driver_4.16.0_all.deb && sudo dpkg -i hailort_4.16.0_RELEVANT_ARCH.deb) and try again?

Regards,