what is the difference between Triton code written for Liger kernels and generated by torch.compile

I'm a beginner. Can you please explain the difference between Triton code written for Liger kernels and Triton code generated by torch.compile? I feel like they seem the same, or maybe I'm mistaken.

The issue names no files, tests, or entry points. Start by reading the repository documentation and comparing its Triton kernel examples with the behavior of torch.compile; done means adding a clear explanation of how the two kinds of Triton code differ.

Written by the indexing model from the issue text.