Having problems with GPU GTX 1650 notebook

Good afternoon.

I had bought an Acer notebook with a GeForce GTX 1650. The notebook OS used to be Linux, but I changed it to Windows 10, bought from MS site.

I’m having AI classes, so my intention buying this notebook was to study and develop AI models.

This way, I started configuring it. Installed CUDA 11 and cuDNN and Driver versions according to this CUDA version.

But Tensorflow and PyTorch was incompatible with this CUDA version. So I had to downgrade it to 10.2 version and cuDNN and Driver versions according to it.

PyTorch worked.

But I was training a CNN model with Tensorflow and it returned the following error message:

UnknownError: Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try looking to see if a warning log message was printed above.
[[node sequential/conv1d/conv1d (defined at :10) ]] [Op:__inference_train_function_3615]

When I looked to prompt window the msg displayed was:

2020-10-05 17:12:17.497166: E tensorflow/stream_executor/cuda/cuda_dnn.cc:329] Could not create cudnn handle: CUDNN_STATUS_NOT_INITIALIZED
2020-10-05 17:12:17.497336: E tensorflow/stream_executor/cuda/cuda_dnn.cc:333] Error retrieving driver version: Unimplemented: kernel reported driver version not implemented on Windows
2020-10-05 17:12:17.500077: E tensorflow/stream_executor/cuda/cuda_dnn.cc:329] Could not create cudnn handle: CUDNN_STATUS_NOT_INITIALIZED
2020-10-05 17:12:17.500245: E tensorflow/stream_executor/cuda/cuda_dnn.cc:333] Error retrieving driver version: Unimplemented: kernel reported driver version not implemented on Windows

Tensorflow version is 2.2.0

There is another issue: when I use Visual Studio, System Info recognizes my GPU but driver verson is assigned to 0.00. When I run a simple code it runs in my CPU not in GPU and shows in “CUDA Overview”:
“No kernel launches captrued! Please verify your activity settings and make sure the kernel filter is setup correctly”.

In the “Session Overview” area it’s shown a warning: “Attempted to perform CUDA profiling on an unsupported device. Kernel launches on such devices are ignored by the profiler”.

How can I get this GPU to work?

Thanks.