Prateek Gupta

@cathyzhyi I have added a PR for this decomposition in functorch: https://github.com/pytorch/functorch/pull/525 So I guess having the decomposition here also will duplicate it.

The [functorch PR]( https://github.com/pytorch/functorch/pull/525) for `aten.native_layer_norm_backward` is merged. Waiting for the integration of latest functorch in torch-mlir. Once done, we can safely close this PR.

Hi @silvasean could you please elaborate on how I can add runtime list type in torch-mlir? It would be great if you could provide some starting point to look at?...

> Adding a runtime list type is quite difficult and not a direction that I think is the best for the project right now. Do you actually have a model...