mirror of https://github.com/pytorch/pytorch.git synced 2025-10-20 21:14:14 +08:00

Files

Ting Lu 297daa1d30 [aarch64] Add sm_80 to CUDA SBSA build (#157843 )

related to https://github.com/pytorch/pytorch/issues/152690

This adds sm_80 to CUDA SBSA builds (12.9), so that we will be able to support Ampere family (e.g: sm_86) and Ada family (e.g: sm_89) on CUDA SBSA builds.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/157843
Approved by: https://github.com/Skylion007, https://github.com/atalman

2025-07-09 11:46:34 +00:00

aarch64_ci_build.sh

[aarch64] Add sm_80 to CUDA SBSA build (#157843 )

2025-07-09 11:46:34 +00:00

aarch64_ci_setup.sh

[CD] Enable python3.13t builds for aarch64 (#144698 )

2025-01-14 02:29:01 +00:00

aarch64_wheel_ci_build.py

Install nvperf_host together with cupti (#156668 )

2025-06-28 04:26:36 +00:00

build_aarch64_wheel.py

Remove outdated instructions from CI scripts (#149795 )

2025-03-22 18:37:07 +00:00

embed_library.py

[CD] Move linux-aarch64 build scripts (#139815 )

2024-11-06 04:16:48 +00:00

README.md

[CD] Move linux-aarch64 build scripts (#139815 )

2024-11-06 04:16:48 +00:00

README.md

Aarch64 (ARM/Graviton) Support Scripts

Scripts for building aarch64 PyTorch PIP Wheels. These scripts build the following wheels:

torch
torchvision
torchaudio
torchtext
torchdata

Aarch64_ci_build.sh

This script is design to support CD operations within PyPi manylinux aarch64 container, and be executed in the container. It prepares the container and then executes aarch64_wheel_ci_build.py to build the wheels. The script "assumes" the PyTorch repo is located at: /pytorch and will put the wheels into /artifacts.

Usage

DESIRED_PYTHON=<PythonVersion> aarch64_ci_build.sh

NOTE: CI build is currently EXPERMINTAL

Build_aarch64_wheel.py

This app allows a person to build using AWS EC3 resources and requires AWS-CLI and Boto3 with AWS credentials to support building EC2 instances for the wheel builds. Can be used in a codebuild CD or from a local system.

Usage

build_aarch64_wheel.py --key-name <YourPemKey> --use-docker --python 3.8 --branch <RCtag>