getting "libpython3.8.so.1.0 cannot open shared object file: No such file or directory" error while running deepstream.py

[ahmedmuzammilAI]

deepstream.py", line 13, in <module>
    import pyds
ImportError: libpython3.8.so.1.0: cannot open shared object file: No such file or directory

can you please help me solve this issue! I have python 3.10 install and i dont know why its trying to find python 3.8

[marcoslucianops]

Probably, you installed a different pyds (deepstream_python_apps) version compared to the DeepStream SDK version.

[ahmedmuzammilAI]

thanks! I uninstall and reinstall pyds using "pip install pyds" command and im getting this error: is there any other way i should install pyds with GPU support? can you please let me know if there is..

Traceback (most recent call last):
  File "/opt/nvidia/deepstream/deepstream-6.4/sources/apps/sample_apps/deepstream-face/DeepStream-Yolo-Face-master/deepstream.py", line 401, in <module>
    sys.exit(main())
  File "/opt/nvidia/deepstream/deepstream-6.4/sources/apps/sample_apps/deepstream-face/DeepStream-Yolo-Face-master/deepstream.py", line 309, in main
    osd.set_property('process-mode', int(pyds.MODE_GPU))
AttributeError: module 'pyds' has no attribute 'MODE_GPU'

[marcoslucianops]

What is your DeepStream version?

[ahmedmuzammilAI]

It's 6.4 Thanks I solved that issue by installing pyds using this command:

wget https://github.com/NVIDIA-AI-IOT/deepstream_python_apps/releases/download/v1.1.10/pyds-1.1.10-py3-none-linux_x86_64.whl

pip3 install pyds-1.1.10-py3-none-linux_x86_64.whl

then I gave the onnx weights and the engine was built successfully but the video gets stuck while opening and it gives segmentation fault(core dumped) error. Please if you could help with this.. is this because of a different deepstream version?

[marcoslucianops]

Are you using the correct exporter file in the utils folder (from this repo)?

[ahmedmuzammilAI]

no sorry! let me check that..

[ahmedmuzammilAI]

Hi @marcoslucianops!! Thanks a lot for your help all along the way you're really awesome! I'm able to run it now and detect faces. I had a few questions if you could please tell me regarding this:

1. will I be able to run this on multiple streams parallelly?
2. can I recognise individual faces with labels/names separately?