Im trying to do some python work based with AI and want to move the stress from cpu to GPU via cuda so my program would run way smoother.

System:
NVIDIA GTX 1650 laptop
I5-9300H
WINDWONS 10 Home 22H2

Executables:
Miniconda3
Microsoft Visual studio 2022 (have tried even the basic VS code)

NVIDIA toolkit version/information

But here I am encountering strange problems:

1. Can’t get any output from pytorch

import torch
torch.cuda.device_count()
torch.cuda.is_available()

Here is the output

2. OpenCV tried manual installation via CMAKE and also premade pip conda install

import cv2
print(cv2.cuda.getCudaEnabledDeviceCount())

In openCV i get output 0

What have i tried:

Tried changing from vs code to microsoft vs code 2022 still same
Tried updating everything in conda: conda update –all