Unable to insert breakpoint in backward(self, grad_output)

[lili.ece.gwu (LII)]

I am writing my own loss function. I find that in version 0.40, the breakpoint doesn’t work in backward(self, grad_output). It is very difficult to debug. Any suggestion, please? Thank you!

[tom (Thomas V)]

It is hard to tell without seeing the function. Are you writing an autograd Function or a loss Module / a python function?
If it is the latter, you could add a backward hook and set a breakpoint in that.

Best regards

Thomas

[lili.ece.gwu (LII)]

Thank you so much for your quick reply, even in Sunday!

Can you give an example? how to use the backward hook?

My original problem is

.backward() not working in a method of a class

Hello everyone. I really need to embed a .backward() Function in a .forward() of another Function, but failed, help! x = np.array([[1.,2.], [3.,4.], [5.,6.]]) x = Variable(torch.from_numpy(x).float(), requires_grad=True) y = var(x) y.backward(y) # it works fine up to here print(x.grad) # but ----------------- class Y_fun(Function): def forward(self, x): y = var(x) y.backward(y) print(x.grad) # exact same thing, but in a forward method, failed …

I appreciate if you could help!