Add LayerNorm layer

[mawanda-jun]

Hi there,
thank you for your groundbreaking work.

I'm trying to develop the rotation equivariant version of this neural network. I succesfully replaced every layer of the network, but there is the [torch.nn.LayerNorm](https://pytorch.org/docs/stable/generated/torch.nn.LayerNorm.html) layer that I wasn't able to replace (I put the InnerBatchNorm instead). Performance lays quite lower wrt its classical version (~15% lower on ImageNet classification) and I think this could be the culprit.

Is there a plan to support this norm layer? I'm willing to contribute, but I think I need guidance.

Cheers.

[psteinb]

I am in the same boat, but with respect to InstanceNorm, see #69

[mawanda-jun]

Definitely interesting, I feel like I need to linger over the theory a bit more to understand how to implement it by myself. Do you think it'd be useful to take a look at our layers together?