That would be nice to have the dot function in pytorch consistent with the numpy one:

For 2-D arrays it is equivalent to matrix multiplication, and for 1-D arrays to inner product of vectors (without complex conjugation). For N dimensions it is a sum product over the last axis of a and the second-to-last of b.
https://docs.scipy.org/doc/numpy/reference/generated/numpy.dot.html

So that with:
x = torch.Tensor(5, 3, 4).normal_()
y = torch.Tensor(4, 7).normal_()

x.dot(y) returns a tensor of shape (5, 3, 7)