Description
🚀 Descirbe the improvement or the new tutorial
PyTorch 2.0 introduced the flagship compilation API, torch.compile, which offers a significant speedup over eager mode execution through graph-level optimization powered by the default TorchInductor backend. While this new feature has generated excitement within the PyTorch community, there is a lack of comprehensive tutorials that delve into the intricacies of torch.compile. The existing tutorials primarily focus on basic usage while missing the essential aspects such as exploring the underlying generated code, debugging potential issues, and conducting performance profiling. Therefore, this proposal aims to address this gap by creating an in-depth tutorial specifically designed for the Inductor CPU backend.
Existing tutorials on this topic
https://pytorch.org/tutorials/intermediate/torch_compile_tutorial.html
Additional context
We aim to complete the document as part of PyTorch Docathon 2023. cc @mingfeima @XiaobingSuper @sanchitintel @ashokei @jingxu10 @ZailiWang @ZhaoqiongZ @leslie-fang-intel @Xia-Weiwen @sekahler2 @CaoE @zhuhaozhe @Valentine233 @EikanWang