Getting weird error in forward pass


Can anyone explain why this is happening? Thanks.

Probably the tensor types (cuda vs cpu or float vs double or so). You could try printing type (input.data) and similar for tge weight.

Best regards Thomas