graph-convnet-tsp
graph-convnet-tsp copied to clipboard
RuntimeError: grad_input must be contiguous
Got error at line loss.backward() in Test backward pass. Do you know the reason?
Are you using the correct PyTorch version (old one)?