For torch.Tesnor it is easy to compute mean and var, but I can not find ways to compute mean and var for Nested Tensors. Nested Tensors support layer_norm operation which include mean and var operation. Is there any ways to compute mean fo Nested Tensors?

Thank for your help!

```
x = torch.randn(1, 192)
y = torch.randn(10, 192)
nested = torch.nested.nested_tensor([x, y])
nested.mean(dim=-1)#Don't support
layer_norm(nested)#support
```