OSSForks/pytorch
1
0
Fork 0
mirror of https://github.com/zebrajr/pytorch.git synced 2025-12-07 00:21:07 +01:00
Code Issues Actions 37 Packages Projects Releases Wiki Activity
51a708ffc6
pytorch/docs/source/torch.compiler_ir.md
Sahdev Zala f34335bf33 Convert compiler rst files to markdown (#155335) 
Convert following compiler rst files to md file.
torch.compiler_inductor_profiling.rst
torch.compiler_ir.rst
torch.compiler_nn_module.rst
torch.compiler_performance_dashboard.rst
torch.compiler_profiling_torch_compile.rst

Fixes #155039

Pull Request resolved: https://github.com/pytorch/pytorch/pull/155335
Approved by: https://github.com/svekars

Co-authored-by: Svetlana Karslioglu <svekars@meta.com>
2025-06-10 01:12:11 +00:00

1.3 KiB
Raw Blame History

IRs

PyTorch 2.0 offers two set of IRs for backends to interface with: Core Aten IR and Prims IR.

Core Aten IR

Core aten ops is the core subset of aten operators that can be used to compose other operators. Core aten IR is fully functional, and there is no inplace or _out variants in this opset. In contrast to Prims IR, core aten ops reuses the existing aten ops in "native_functions.yaml", and it doesn't further decompose ops into explicit type promotion and broadcasting ops. This opset is designed to serve as the functional IR to interface with backends.

  This opset is still under active development, more ops will be added in the future.

   :file: ../build/ir/aten_ops.csv
   :widths: auto
   :header-rows: 1

Prims IR

Prims IR is a set of primitive operators that can be used to compose other operators. Prims IR is a lower level opset than core aten IR, and it further decomposes ops into explicit type promotion and broadcasting ops: prims.convert_element_type and prims.broadcast_in_dim. This opset is designed to interface with compiler backends.

  This opset is still under active development, more ops will be added in the future.

   :file: ../build/ir/prims_ops.csv
   :widths: auto
   :header-rows: 1