How to turn a list of tensor to tensor?

To me it’s sort of unintuitive, why wouldn’t using the tensor class work?

torch.tensor([ torch.tensor([i]).repeat(15) for i in range(0,5)])

the list is the same size and it’s really a matrix/tensor already…but somehow only:

torch.stack([ torch.tensor([i]).repeat(15) for i in range(0,5)])
tensor([[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
        [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
        [2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2],
        [3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3],
        [4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4]])

worked.

Btw, is this this most efficient way to do it in a vectorized way?

3 Likes