I think this is another case in which the torch version that the grader is using is less permissive than the one in the notebook. It looks like the grader’s version of torch.cat can’t handle different types. I wrote that code like this and it passes the grader:
combined = torch.cat((x.float(), y.float()), dim = 1)
The point is that I do the conversion to float before the cat, of course.