How to interpret Tensor sizes?

Question

I\'m having difficulty understanding the difference between the following:

x1 = torch.tensor([1, 2, 3]) # single brackets