Say I have a 3-dim tensor (A) and 2 long tensor (B and C, with only 1 dimension) to be used for indexing. I want to index all of the first dimension and the last 2 dimension with B and C.

In PyTorch, I can index tensor A with `A[:, B, C]`

:

```
>>> A
tensor([[[ 0, 1, 2, 3],
[ 4, 5, 6, 7],
[ 8, 9, 10, 11]],
[[12, 13, 14, 15],
[16, 17, 18, 19],
[20, 21, 22, 23]]])
>>> B
tensor([0, 2])
>>> C
tensor([3, 1])
>>> A[:, B, C]
tensor([[ 3, 9],
[15, 21]])
```

How can I do the equivalent in libtorch? I know I can use `A.index({B, C, D});`

if I want to index all dimensions with B, C, D, but how can I filter only the last two dimension with B and C?