Get error message when excute "loss.backward()"

I’m learning “Neural Transfer Using Pytorch” example, I have typed the same code with the Tutorials, but I get the error, anyone can help me?

Error Message

RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.cuda.FloatTensor [128, 65536]], which is output 0 of ViewBackward, is at version 2; expected version 1 instead. Hint: the backtrace further above shows the operation that failed to compute its gradient. The variable in question was changed in there or anywhere later.

It is quite impossible to track down the issue without any code/example. Can you please share some working codes to replicate the issue?