Python neural network with Keras runs on CPU, but crashes on the GPU

I implemented a neural network that learns to play PacMan using gym,box2d and gym[atari] with Keras models. The training was very slow so I tried to make in run on my GTX 1060 Max-Q. I installed the latest version of Tensorflow, installed CUDA 11.0 and cuDNN 8.0.4.30. The program opens all the libraries succesfully, detects the GPU correctly, creates the Tensor device, starts the first frame of the render, freezes for about 9 seconds and then exits with code -1073740791 (0xC0000409).

Why is this happening and what can I do to fix it?

-1073740791 (0xC0000409) is a stack buffer overflow on windows machines

Here's some documentation for it.

You need to make the training file smaller or run it on a better PC