PyTorch cuDNN Convolution

PyTorch extension enabling direct access to the following cuDNN-accelerated C++ functions that are included in PyTorch:

• cudnn_convolution
• cudnn_convolution_backward_weight
• cudnn_convolution_backward_input

The functions defined here can be called from Python in replacement of torch.nn.conv2d, torch.nn.grad.conv2d_weight and torch.nn.grad.conv2d_input, and run significantly faster. See example.py for how these functions are called.

Adapted from the following code posted by hanspinckaers:

https://discuss.pytorch.org/t/cuda-error-with-cudnn-convolution-backward-weight-function/41214