Tensorflow and dependent versions compatibility (cuda, cuDNN) and GPU detection issue

AI & Data Science Deep Learning (Training & Inference) TensorRT
, ,
1

I need to run a model in the tensorflow library. My GPU supports up to version 2.10.0. I chose to use this version (the latest that supports it). To do this, I installed CUDA and cuDNN in the appropriate versions as I saw here:

The problem is that tensorflow does not recognize my GPU.

Environment
TensorFlow version (if applicable): 2.10.0
GPU type: NVIDIA GeForce RTX 4050 laptop GPU
Nvidia driver version:
CUDA version: 11.8
CUDNN version: 8.6.0
OS + version: windows 11, 64
Python version (if applicable): Python 3.9.13

I searched for existing answers and I can’t find a solution to the problem.
I would appreciate help on this topic.
Thank you very much!!

2

I see that there was something like this that he was able to solve it: TensorFlow Can't Detect GPU (4090) with cuDNN 8.6 and CUDA 11.8 Installed - Stack Overflow
But I didn’t understand how.
I would be happy to help.