Does it remove the old Dispatch.h header?

No, it adds the new Dispatch_v2.h header while keeping the old one to ensure other legacy code in the file continues to function without breaking.

What is the main benefit of using AT_DISPATCH_V2 over the legacy macros?

AT_DISPATCH_V2 eliminates the need for arity-specific macro names (like AND2, AND3) and allows for more composable, extensible type sets using AT_EXPAND.

Which files are best suited for this skill?

It is ideal for PyTorch native operator implementations, ATen kernel files, and CUDA kernels typically located within the aten/src/ATen/ directory.

Does this skill handle complex lambdas with commas?

Yes, it automatically wraps lambdas in the mandatory AT_WRAP macro to prevent C++ preprocessor parsing issues when the lambda contains internal commas.

PyTorch AT_DISPATCH V2 Converter

Name: PyTorch AT_DISPATCH V2 Converter
Author: Activer007

byActiver007

开发者工具

Refactors legacy PyTorch AT_DISPATCH macros into the modern AT_DISPATCH_V2 format for ATen C++ development.

关于

This skill automates the migration of legacy PyTorch dispatch macros to the updated V2 API found in ATen. It streamlines the process of porting ATen kernels and native operators by correctly reordering arguments, wrapping lambdas with the required AT_WRAP macro, and expanding type groups using AT_EXPAND. It is specifically designed for developers working on PyTorch core, CUDA kernels, or custom operators who need to maintain compatibility with the latest ATen dispatching standards while reducing boilerplate and macro arity issues.

主要功能

0 GitHub stars
Automatic header injection for ATen/Dispatch_v2.h
Support for complex multi-line lambdas with internal commas
Automatic argument reordering for AT_DISPATCH_V2 compliance
Conversion of implicit macro type sets to explicit AT_EXPAND type groups
Safe lambda wrapping using AT_WRAP to prevent macro parsing errors

使用场景

Refactoring ATen native operators to use composable type sets
Porting legacy PyTorch C++ kernels to the modern Dispatch V2 API
Upgrading custom CUDA kernels to follow current PyTorch coding standards

关于

主要功能

0 GitHub stars
Automatic header injection for ATen/Dispatch_v2.h
Support for complex multi-line lambdas with internal commas
Automatic argument reordering for AT_DISPATCH_V2 compliance
Conversion of implicit macro type sets to explicit AT_EXPAND type groups
Safe lambda wrapping using AT_WRAP to prevent macro parsing errors

使用场景

Refactoring ATen native operators to use composable type sets
Porting legacy PyTorch C++ kernels to the modern Dispatch V2 API
Upgrading custom CUDA kernels to follow current PyTorch coding standards