I am working on LSTMs and I want to compute cross_entropy loss, given X and y.

X.shape: (batch_size, time_steps, number_of_classes)

y.shape: (batch_size, time_steps)

y contains the ground truth indices, such as 0,1,2,3,..,10.

When I use torch.nn.functional.cross_entropy(X, y)

I get error:

ValueError: Expected target size (1000, 10), got torch.Size([1000, 1])

What is the correct way to use PyTorch's cross_entropy with this kind of Sequential model?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.