2D tensors should be reversed?

Hi,
When implementing a linear layer, the following code does not produce the expected output:
```Julia
import Torch: tensor
W = collect(reshape(1.f0:6.f0, (3,2)))
x = reshape([1.f0; 1.f0], (2, 1))
expected = W * x
output = tensor(W, dev = 0) * tensor(x, dev = 0)
```
I noted that 2d and nd tensors are treated differently and to me this looks like the root cause:
https://github.com/FluxML/Torch.jl/blob/master/src/tensor.jl#L143-L149

Is there a good reason to treat them differently?

Reproduce the Julia example in the issue and compare tensor(W) * tensor(x) with W * x. Read src/tensor.jl around lines 143-149, where 2D and higher-dimensional tensors are treated differently. Done means determining whether that distinction causes the mismatch and documenting or correcting the behavior so the tensor result matches the stated expected output.

Written by the indexing model from the issue text.