A RuntimeError

[ReunionJ]

Hello, when I run your code, I encountered the following problems when calculating the loss of classifier:
"RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.cuda.FloatTensor [256, 4]], which is output 0 of TBackward, 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. Good luck!"

[Jcxloyal]

hello friend, I met the same problem, I solved it by rearranging the code in the model.py

hope it helps

[ReunionJ]

hello friend, I met the same problem, I solved it by rearranging the code in the model.py

hope it helps

Thank you very much for your reply!
But after modifying the code, I encountered the situation that the accuracy of model prediction is very low. I don't know if it is because the epoch is too small.
Do you encounter the same situation?

[Jcxloyal]

hello friend, I met the same problem, I solved it by rearranging the code in the model.py
hope it helps

Thank you very much for your reply!
But after modifying the code, I encountered the situation that the accuracy of model prediction is very low. I don't know if it is because the epoch is too small.
Do you encounter the same situation?

I'm sorry, friend. I didn't encounter the situation

[sunhufei]

您好，这个程序可以把你改过的代码发给我一份ma？