Issues using visual studio code & pytorch

Hi all,
I’m new using jetson nano, and i’m trying to run an object detection model from yolov8. I have succesfully run it on a PC with CUDA, but i’ve encountered numerous issues, one of wich involves the communication between CUDA and the virtual enviroment. This issue prevents my object detector from running as it should
P. S. I’m not using conda

Thanks a lot for the help