[CI] Delete binary smoke workflows (#164260)

Those were very useful in the past, because:
- CI builder jobs did not generates wheels, but rather run `python setup.py develop` and shared docker layers, which is no longer the case, all CI jobs produce wheels
- CD jobs were targeting pre-CXX11 ABI, but this is no longer the case after manylinux2_28 migration

Existing, but acceptable gaps:
 - Windows libtorch debug builds sometimes might fail, but IMO it's ok not to be able to produce those for a few days, as number of libtorch users are somewhat small
 - All CD jobs are based on AlmaLinux, while CI are based on Ubuntu, but this could be adjusted if needed, besides AlmaLinux-9 and Ubuntu-22.04 are pretty close in terms of glibc and gcc versions
 - CD jobs build for all GPU architectures, while CI only for the one being tested, but there are now periodic H100 and B200 jobs, and not a lot of development happens for Voltas or Pascals

Besides there are better tools to alert about the nightly failures

Pull Request resolved: https://github.com/pytorch/pytorch/pull/164260
Approved by: https://github.com/seemethere, https://github.com/atalman
This commit is contained in:
Nikita Shulga
2025-09-30 10:32:46 -07:00
committed by PyTorch MergeBot
parent e30f01b5b5
commit 5a93f00c79
6 changed files with 0 additions and 926 deletions

View File

@ -127,53 +127,6 @@ LINUX_BINARY_BUILD_WORFKLOWS = [
),
]
ROCM_SMOKE_WORKFLOWS = [
BinaryBuildWorkflow(
os=OperatingSystem.LINUX,
package_type="manywheel",
build_variant="rocm",
build_configs=generate_binary_build_matrix.generate_wheels_matrix(
OperatingSystem.LINUX,
arches=["6.4"],
python_versions=["3.10"],
),
ciflow_config=CIFlowConfig(
labels={
LABEL_CIFLOW_BINARIES,
LABEL_CIFLOW_BINARIES_WHEEL,
LABEL_CIFLOW_ROCM,
},
isolated_workflow=True,
),
branches="main",
),
]
LINUX_BINARY_SMOKE_WORKFLOWS = [
BinaryBuildWorkflow(
os=OperatingSystem.LINUX,
package_type="manywheel",
build_configs=generate_binary_build_matrix.generate_wheels_matrix(
OperatingSystem.LINUX,
arches=["13.0"],
python_versions=["3.12"],
),
branches="main",
),
BinaryBuildWorkflow(
os=OperatingSystem.LINUX,
package_type="libtorch",
build_variant=generate_binary_build_matrix.RELEASE,
build_configs=generate_binary_build_matrix.generate_libtorch_matrix(
OperatingSystem.LINUX,
generate_binary_build_matrix.RELEASE,
arches=["cpu"],
libtorch_variants=["shared-with-deps"],
),
branches="main",
),
]
WINDOWS_BINARY_BUILD_WORKFLOWS = [
BinaryBuildWorkflow(
os=OperatingSystem.WINDOWS,
@ -259,39 +212,6 @@ WINDOWS_BINARY_BUILD_WORKFLOWS = [
),
]
WINDOWS_BINARY_SMOKE_WORKFLOWS = [
BinaryBuildWorkflow(
os=OperatingSystem.WINDOWS,
package_type="libtorch",
build_variant=generate_binary_build_matrix.RELEASE,
build_configs=generate_binary_build_matrix.generate_libtorch_matrix(
OperatingSystem.WINDOWS,
generate_binary_build_matrix.RELEASE,
arches=["cpu"],
libtorch_variants=["shared-with-deps"],
),
branches="main",
ciflow_config=CIFlowConfig(
isolated_workflow=True,
),
),
BinaryBuildWorkflow(
os=OperatingSystem.WINDOWS,
package_type="libtorch",
build_variant=generate_binary_build_matrix.DEBUG,
build_configs=generate_binary_build_matrix.generate_libtorch_matrix(
OperatingSystem.WINDOWS,
generate_binary_build_matrix.DEBUG,
arches=["cpu"],
libtorch_variants=["shared-with-deps"],
),
branches="main",
ciflow_config=CIFlowConfig(
isolated_workflow=True,
),
),
]
MACOS_BINARY_BUILD_WORKFLOWS = [
BinaryBuildWorkflow(
os=OperatingSystem.MACOS_ARM64,
@ -372,23 +292,10 @@ def main() -> None:
jinja_env.get_template("linux_binary_build_workflow.yml.j2"),
S390X_BINARY_BUILD_WORKFLOWS,
),
(
# Give rocm it's own workflow file
jinja_env.get_template("linux_binary_build_workflow.yml.j2"),
ROCM_SMOKE_WORKFLOWS,
),
(
jinja_env.get_template("linux_binary_build_workflow.yml.j2"),
LINUX_BINARY_SMOKE_WORKFLOWS,
),
(
jinja_env.get_template("windows_binary_build_workflow.yml.j2"),
WINDOWS_BINARY_BUILD_WORKFLOWS,
),
(
jinja_env.get_template("windows_binary_build_workflow.yml.j2"),
WINDOWS_BINARY_SMOKE_WORKFLOWS,
),
(
jinja_env.get_template("macos_binary_build_workflow.yml.j2"),
MACOS_BINARY_BUILD_WORKFLOWS,

View File

@ -1,87 +0,0 @@
# @generated DO NOT EDIT MANUALLY
# Template is at: .github/templates/linux_binary_build_workflow.yml.j2
# Generation script: .github/scripts/generate_ci_workflows.py
name: linux-binary-libtorch-release
on:
push:
branches:
- main
tags:
- 'ciflow/trunk/*'
workflow_dispatch:
permissions:
id-token: write
env:
# Needed for conda builds
ALPINE_IMAGE: "308535385114.dkr.ecr.us-east-1.amazonaws.com/tool/alpine"
AWS_DEFAULT_REGION: us-east-1
BINARY_ENV_FILE: /tmp/env
BUILD_ENVIRONMENT: linux-binary-libtorch-release
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
PR_NUMBER: ${{ github.event.pull_request.number }}
PYTORCH_FINAL_PACKAGE_DIR: /artifacts
PYTORCH_ROOT: /pytorch
SHA1: ${{ github.event.pull_request.head.sha || github.sha }}
SKIP_ALL_TESTS: 0
concurrency:
group: linux-binary-libtorch-release-${{ github.event.pull_request.number || github.ref_name }}-${{ github.ref_type == 'branch' && github.sha }}-${{ github.event_name == 'workflow_dispatch' }}
cancel-in-progress: true
jobs:
get-label-type:
if: github.repository_owner == 'pytorch'
name: get-label-type
uses: pytorch/pytorch/.github/workflows/_runner-determinator.yml@main
with:
triggering_actor: ${{ github.triggering_actor }}
issue_owner: ${{ github.event.pull_request.user.login || github.event.issue.user.login }}
curr_branch: ${{ github.head_ref || github.ref_name }}
curr_ref_type: ${{ github.ref_type }}
libtorch-cpu-shared-with-deps-release-build:
if: ${{ github.repository_owner == 'pytorch' }}
uses: ./.github/workflows/_binary-build-linux.yml
needs: get-label-type
with:
PYTORCH_ROOT: /pytorch
PACKAGE_TYPE: libtorch
# TODO: This is a legacy variable that we eventually want to get rid of in
# favor of GPU_ARCH_VERSION
DESIRED_CUDA: cpu
GPU_ARCH_TYPE: cpu
DOCKER_IMAGE: libtorch-cxx11-builder
DOCKER_IMAGE_TAG_PREFIX: cpu
LIBTORCH_CONFIG: release
LIBTORCH_VARIANT: shared-with-deps
runner_prefix: "${{ needs.get-label-type.outputs.label-type }}"
build_name: libtorch-cpu-shared-with-deps-release
build_environment: linux-binary-libtorch-release
secrets:
github-token: ${{ secrets.GITHUB_TOKEN }}
libtorch-cpu-shared-with-deps-release-test: # Testing
if: ${{ github.repository_owner == 'pytorch' }}
needs:
- libtorch-cpu-shared-with-deps-release-build
- get-label-type
uses: ./.github/workflows/_binary-test-linux.yml
with:
PYTORCH_ROOT: /pytorch
PACKAGE_TYPE: libtorch
# TODO: This is a legacy variable that we eventually want to get rid of in
# favor of GPU_ARCH_VERSION
DESIRED_CUDA: cpu
GPU_ARCH_TYPE: cpu
DOCKER_IMAGE: libtorch-cxx11-builder
DOCKER_IMAGE_TAG_PREFIX: cpu
LIBTORCH_CONFIG: release
LIBTORCH_VARIANT: shared-with-deps
build_name: libtorch-cpu-shared-with-deps-release
build_environment: linux-binary-libtorch-release
runner_prefix: "${{ needs.get-label-type.outputs.label-type }}"
runs_on: linux.4xlarge
secrets:
github-token: ${{ secrets.GITHUB_TOKEN }}

View File

@ -1,88 +0,0 @@
# @generated DO NOT EDIT MANUALLY
# Template is at: .github/templates/linux_binary_build_workflow.yml.j2
# Generation script: .github/scripts/generate_ci_workflows.py
name: linux-binary-manywheel
on:
push:
branches:
- main
tags:
- 'ciflow/trunk/*'
workflow_dispatch:
permissions:
id-token: write
env:
# Needed for conda builds
ALPINE_IMAGE: "308535385114.dkr.ecr.us-east-1.amazonaws.com/tool/alpine"
AWS_DEFAULT_REGION: us-east-1
BINARY_ENV_FILE: /tmp/env
BUILD_ENVIRONMENT: linux-binary-manywheel
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
PR_NUMBER: ${{ github.event.pull_request.number }}
PYTORCH_FINAL_PACKAGE_DIR: /artifacts
PYTORCH_ROOT: /pytorch
SHA1: ${{ github.event.pull_request.head.sha || github.sha }}
SKIP_ALL_TESTS: 0
concurrency:
group: linux-binary-manywheel-${{ github.event.pull_request.number || github.ref_name }}-${{ github.ref_type == 'branch' && github.sha }}-${{ github.event_name == 'workflow_dispatch' }}
cancel-in-progress: true
jobs:
get-label-type:
if: github.repository_owner == 'pytorch'
name: get-label-type
uses: pytorch/pytorch/.github/workflows/_runner-determinator.yml@main
with:
triggering_actor: ${{ github.triggering_actor }}
issue_owner: ${{ github.event.pull_request.user.login || github.event.issue.user.login }}
curr_branch: ${{ github.head_ref || github.ref_name }}
curr_ref_type: ${{ github.ref_type }}
manywheel-py3_12-cuda13_0-build:
if: ${{ github.repository_owner == 'pytorch' }}
uses: ./.github/workflows/_binary-build-linux.yml
needs: get-label-type
with:
PYTORCH_ROOT: /pytorch
PACKAGE_TYPE: manywheel
# TODO: This is a legacy variable that we eventually want to get rid of in
# favor of GPU_ARCH_VERSION
DESIRED_CUDA: cu130
GPU_ARCH_VERSION: "13.0"
GPU_ARCH_TYPE: cuda
DOCKER_IMAGE: manylinux2_28-builder
DOCKER_IMAGE_TAG_PREFIX: cuda13.0
DESIRED_PYTHON: "3.12"
runner_prefix: "${{ needs.get-label-type.outputs.label-type }}"
build_name: manywheel-py3_12-cuda13_0
build_environment: linux-binary-manywheel
PYTORCH_EXTRA_INSTALL_REQUIREMENTS: nvidia-cuda-nvrtc==13.0.48; platform_system == 'Linux' | nvidia-cuda-runtime==13.0.48; platform_system == 'Linux' | nvidia-cuda-cupti==13.0.48; platform_system == 'Linux' | nvidia-cudnn-cu13==9.13.0.50; platform_system == 'Linux' | nvidia-cublas==13.0.0.19; platform_system == 'Linux' | nvidia-cufft==12.0.0.15; platform_system == 'Linux' | nvidia-curand==10.4.0.35; platform_system == 'Linux' | nvidia-cusolver==12.0.3.29; platform_system == 'Linux' | nvidia-cusparse==12.6.2.49; platform_system == 'Linux' | nvidia-cusparselt-cu13==0.8.0; platform_system == 'Linux' | nvidia-nccl-cu13==2.28.3; platform_system == 'Linux' | nvidia-nvshmem-cu13==3.3.24; platform_system == 'Linux' | nvidia-nvtx==13.0.39; platform_system == 'Linux' | nvidia-nvjitlink==13.0.39; platform_system == 'Linux' | nvidia-cufile==1.15.0.42; platform_system == 'Linux'
secrets:
github-token: ${{ secrets.GITHUB_TOKEN }}
manywheel-py3_12-cuda13_0-test: # Testing
if: ${{ github.repository_owner == 'pytorch' }}
needs:
- manywheel-py3_12-cuda13_0-build
- get-label-type
uses: ./.github/workflows/_binary-test-linux.yml
with:
PYTORCH_ROOT: /pytorch
PACKAGE_TYPE: manywheel
# TODO: This is a legacy variable that we eventually want to get rid of in
# favor of GPU_ARCH_VERSION
DESIRED_CUDA: cu130
GPU_ARCH_VERSION: "13.0"
GPU_ARCH_TYPE: cuda
DOCKER_IMAGE: manylinux2_28-builder
DOCKER_IMAGE_TAG_PREFIX: cuda13.0
DESIRED_PYTHON: "3.12"
build_name: manywheel-py3_12-cuda13_0
build_environment: linux-binary-manywheel
runner_prefix: "${{ needs.get-label-type.outputs.label-type }}"
runs_on: linux.g4dn.4xlarge.nvidia.gpu # 12.8+ builds need sm_70+ runner
secrets:
github-token: ${{ secrets.GITHUB_TOKEN }}

View File

@ -1,136 +0,0 @@
# @generated DO NOT EDIT MANUALLY
# Template is at: .github/templates/linux_binary_build_workflow.yml.j2
# Generation script: .github/scripts/generate_ci_workflows.py
name: linux-binary-manywheel-rocm
on:
push:
branches:
- main
tags:
- 'ciflow/binaries/*'
- 'ciflow/binaries_wheel/*'
- 'ciflow/rocm/*'
workflow_dispatch:
permissions:
id-token: write
env:
# Needed for conda builds
ALPINE_IMAGE: "308535385114.dkr.ecr.us-east-1.amazonaws.com/tool/alpine"
AWS_DEFAULT_REGION: us-east-1
BINARY_ENV_FILE: /tmp/env
BUILD_ENVIRONMENT: linux-binary-manywheel-rocm
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
PR_NUMBER: ${{ github.event.pull_request.number }}
PYTORCH_FINAL_PACKAGE_DIR: /artifacts
PYTORCH_ROOT: /pytorch
SHA1: ${{ github.event.pull_request.head.sha || github.sha }}
SKIP_ALL_TESTS: 0
concurrency:
group: linux-binary-manywheel-rocm-${{ github.event.pull_request.number || github.ref_name }}-${{ github.ref_type == 'branch' && github.sha }}-${{ github.event_name == 'workflow_dispatch' }}
cancel-in-progress: true
jobs:
get-label-type:
if: github.repository_owner == 'pytorch'
name: get-label-type
uses: pytorch/pytorch/.github/workflows/_runner-determinator.yml@main
with:
triggering_actor: ${{ github.triggering_actor }}
issue_owner: ${{ github.event.pull_request.user.login || github.event.issue.user.login }}
curr_branch: ${{ github.head_ref || github.ref_name }}
curr_ref_type: ${{ github.ref_type }}
manywheel-py3_10-rocm6_4-build:
if: ${{ github.repository_owner == 'pytorch' }}
uses: ./.github/workflows/_binary-build-linux.yml
needs: get-label-type
with:
PYTORCH_ROOT: /pytorch
PACKAGE_TYPE: manywheel
# TODO: This is a legacy variable that we eventually want to get rid of in
# favor of GPU_ARCH_VERSION
DESIRED_CUDA: rocm6.4
GPU_ARCH_VERSION: "6.4"
GPU_ARCH_TYPE: rocm
DOCKER_IMAGE: manylinux2_28-builder
DOCKER_IMAGE_TAG_PREFIX: rocm6.4
DESIRED_PYTHON: "3.10"
runner_prefix: "${{ needs.get-label-type.outputs.label-type }}"
timeout-minutes: 300
build_name: manywheel-py3_10-rocm6_4
build_environment: linux-binary-manywheel-rocm
secrets:
github-token: ${{ secrets.GITHUB_TOKEN }}
manywheel-py3_10-rocm6_4-test: # Testing
if: ${{ github.repository_owner == 'pytorch' }}
needs:
- manywheel-py3_10-rocm6_4-build
- get-label-type
runs-on: linux.rocm.gpu.mi250
timeout-minutes: 240
env:
PYTORCH_ROOT: /pytorch
PACKAGE_TYPE: manywheel
# TODO: This is a legacy variable that we eventually want to get rid of in
# favor of GPU_ARCH_VERSION
DESIRED_CUDA: rocm6.4
GPU_ARCH_VERSION: "6.4"
GPU_ARCH_TYPE: rocm
SKIP_ALL_TESTS: 1
DOCKER_IMAGE: manylinux2_28-builder
DOCKER_IMAGE_TAG_PREFIX: rocm6.4
DESIRED_PYTHON: "3.10"
steps:
- name: Setup ROCm
uses: ./.github/actions/setup-rocm
- uses: actions/download-artifact@v4.1.7
name: Download Build Artifacts
with:
name: manywheel-py3_10-rocm6_4
path: "${{ runner.temp }}/artifacts/"
- name: Checkout PyTorch
uses: actions/checkout@v4
with:
ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
submodules: recursive
path: pytorch
show-progress: false
- name: Clean PyTorch checkout
run: |
# Remove any artifacts from the previous checkouts
git clean -fxd
working-directory: pytorch
- name: ROCm set GPU_FLAG
run: |
echo "GPU_FLAG=--device=/dev/mem --device=/dev/kfd --device=/dev/dri --group-add video --group-add daemon" >> "${GITHUB_ENV}"
- name: configure aws credentials
id: aws_creds
if: ${{ startsWith(github.event.ref, 'refs/tags/ciflow/') }}
uses: aws-actions/configure-aws-credentials@v4
with:
role-to-assume: arn:aws:iam::308535385114:role/gha_workflow_s3_and_ecr_read_only
aws-region: us-east-1
role-duration-seconds: 18000
- name: Calculate docker image
id: calculate-docker-image
uses: pytorch/test-infra/.github/actions/calculate-docker-image@main
with:
docker-registry: ${{ startsWith(github.event.ref, 'refs/tags/ciflow/') && '308535385114.dkr.ecr.us-east-1.amazonaws.com' || 'docker.io' }}
docker-image-name: manylinux2_28-builder
custom-tag-prefix: rocm6.4
docker-build-dir: .ci/docker
working-directory: pytorch
- name: Pull Docker image
uses: pytorch/test-infra/.github/actions/pull-docker-image@main
with:
docker-image: ${{ steps.calculate-docker-image.outputs.docker-image }}
- name: Test Pytorch binary
uses: ./pytorch/.github/actions/test-pytorch-binary
env:
DOCKER_IMAGE: ${{ steps.calculate-docker-image.outputs.docker-image }}
- name: Teardown ROCm
uses: ./.github/actions/teardown-rocm

View File

@ -1,261 +0,0 @@
# @generated DO NOT EDIT MANUALLY
# Template is at: .github/templates/windows_binary_build_workflow.yml.j2
# Generation script: .github/scripts/generate_ci_workflows.py
name: windows-binary-libtorch-debug
on:
push:
branches:
- main
workflow_dispatch:
env:
# Needed for conda builds
ALPINE_IMAGE: "308535385114.dkr.ecr.us-east-1.amazonaws.com/tool/alpine"
AWS_DEFAULT_REGION: us-east-1
BUILD_ENVIRONMENT: windows-binary-libtorch-debug
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
PR_NUMBER: ${{ github.event.pull_request.number }}
SHA1: ${{ github.event.pull_request.head.sha || github.sha }}
SKIP_ALL_TESTS: 1
OS: windows
concurrency:
group: windows-binary-libtorch-debug-${{ github.event.pull_request.number || github.ref_name }}-${{ github.ref_type == 'branch' && github.sha }}-${{ github.event_name == 'workflow_dispatch' }}
cancel-in-progress: true
jobs:
get-label-type:
if: github.repository_owner == 'pytorch'
name: get-label-type
uses: pytorch/pytorch/.github/workflows/_runner-determinator.yml@main
with:
triggering_actor: ${{ github.triggering_actor }}
issue_owner: ${{ github.event.pull_request.user.login || github.event.issue.user.login }}
curr_branch: ${{ github.head_ref || github.ref_name }}
curr_ref_type: ${{ github.ref_type }}
libtorch-cpu-shared-with-deps-debug-build:
if: ${{ github.repository_owner == 'pytorch' }}
needs: get-label-type
runs-on: "${{ needs.get-label-type.outputs.label-type }}windows.4xlarge.nonephemeral"
timeout-minutes: 360
env:
PYTORCH_ROOT: ${{ github.workspace }}/pytorch
PACKAGE_TYPE: libtorch
# TODO: This is a legacy variable that we eventually want to get rid of in
# favor of GPU_ARCH_VERSION
DESIRED_CUDA: cpu
GPU_ARCH_TYPE: cpu
SKIP_ALL_TESTS: 1
LIBTORCH_CONFIG: debug
LIBTORCH_VARIANT: shared-with-deps
# This is a dummy value for libtorch to work correctly with our batch scripts
# without this value pip does not get installed for some reason
DESIRED_PYTHON: "3.10"
steps:
# NOTE: These environment variables are put here so that they can be applied on every job equally
# They are also here because setting them at a workflow level doesn't give us access to the
# runner.temp variable, which we need.
- name: Populate binary env
shell: bash
run: |
echo "BINARY_ENV_FILE=${RUNNER_TEMP}/env" >> "${GITHUB_ENV}"
echo "PYTORCH_FINAL_PACKAGE_DIR=${RUNNER_TEMP}/artifacts" >> "${GITHUB_ENV}"
echo "WIN_PACKAGE_WORK_DIR=${RUNNER_TEMP}"
- name: Display EC2 information
shell: bash
run: |
set -euo pipefail
function get_ec2_metadata() {
# Pulled from instance metadata endpoint for EC2
# see https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/instancedata-data-retrieval.html
category=$1
curl -H "X-aws-ec2-metadata-token: $(curl -s -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-ec2-metadata-token-ttl-seconds: 30")" -fsSL "http://169.254.169.254/latest/meta-data/${category}"
}
echo "ami-id: $(get_ec2_metadata ami-id)"
echo "instance-id: $(get_ec2_metadata instance-id)"
echo "instance-type: $(get_ec2_metadata instance-type)"
echo "system info $(uname -a)"
- name: "[FB EMPLOYEES] Enable SSH (Click me for login details)"
uses: pytorch/test-infra/.github/actions/setup-ssh@main
continue-on-error: true
with:
github-secret: ${{ secrets.GITHUB_TOKEN }}
- name: Enable git long paths and symlinks on Windows and disable fsmonitor daemon
shell: bash
run: |
git config --global core.longpaths true
git config --global core.symlinks true
# https://git-scm.com/docs/git-fsmonitor--daemon. The daemon could lock
# the directory on Windows and prevent GHA from checking out as reported
# in https://github.com/actions/checkout/issues/1018
git config --global core.fsmonitor false
# Needed for binary builds, see: https://github.com/pytorch/pytorch/issues/73339#issuecomment-1058981560
- name: Enable long paths on Windows
shell: powershell
run: |
Set-ItemProperty -Path "HKLM:\\SYSTEM\CurrentControlSet\Control\FileSystem" -Name "LongPathsEnabled" -Value 1
# Since it's just a defensive command, the workflow should continue even the command fails. This step can be
# removed once Windows Defender is removed from the AMI
- name: Disables Windows Defender scheduled and real-time scanning for files in directories used by PyTorch
continue-on-error: true
shell: powershell
run: |
Add-MpPreference -ExclusionPath $(Get-Location).tostring(),$Env:TEMP -ErrorAction Ignore
# Let's both exclude the path and disable Windows Defender completely just to be sure
# that it doesn't interfere
Set-MpPreference -DisableRealtimeMonitoring $True -ErrorAction Ignore
- name: Checkout PyTorch
uses: actions/checkout@v4
with:
ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
submodules: recursive
path: pytorch
show-progress: false
- name: Clean PyTorch checkout
run: |
# Remove any artifacts from the previous checkouts
git clean -fxd
working-directory: pytorch
- name: Populate binary env
shell: bash
run: |
"${PYTORCH_ROOT}/.circleci/scripts/binary_populate_env.sh"
- name: Build PyTorch binary
shell: bash
run: |
"${PYTORCH_ROOT}/.circleci/scripts/binary_windows_build.sh"
- uses: actions/upload-artifact@v4.4.0
if: always()
with:
name: libtorch-cpu-shared-with-deps-debug
retention-days: 14
if-no-files-found: error
path: "${{ env.PYTORCH_FINAL_PACKAGE_DIR }}"
- name: Wait until all sessions have drained
shell: powershell
working-directory: pytorch
if: always()
timeout-minutes: 120
run: |
.github\scripts\wait_for_ssh_to_drain.ps1
- name: Kill active ssh sessions if still around (Useful if workflow was cancelled)
shell: powershell
working-directory: pytorch
if: always()
run: |
.github\scripts\kill_active_ssh_sessions.ps1
libtorch-cpu-shared-with-deps-debug-test: # Testing
if: ${{ github.repository_owner == 'pytorch' }}
needs:
- libtorch-cpu-shared-with-deps-debug-build
- get-label-type
runs-on: "${{ needs.get-label-type.outputs.label-type }}windows.4xlarge.nonephemeral"
timeout-minutes: 360
env:
PYTORCH_ROOT: ${{ github.workspace }}/pytorch
PACKAGE_TYPE: libtorch
# TODO: This is a legacy variable that we eventually want to get rid of in
# favor of GPU_ARCH_VERSION
DESIRED_CUDA: cpu
GPU_ARCH_TYPE: cpu
SKIP_ALL_TESTS: 1
LIBTORCH_CONFIG: debug
LIBTORCH_VARIANT: shared-with-deps
# This is a dummy value for libtorch to work correctly with our batch scripts
# without this value pip does not get installed for some reason
DESIRED_PYTHON: "3.10"
steps:
- name: Display EC2 information
shell: bash
run: |
set -euo pipefail
function get_ec2_metadata() {
# Pulled from instance metadata endpoint for EC2
# see https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/instancedata-data-retrieval.html
category=$1
curl -H "X-aws-ec2-metadata-token: $(curl -s -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-ec2-metadata-token-ttl-seconds: 30")" -fsSL "http://169.254.169.254/latest/meta-data/${category}"
}
echo "ami-id: $(get_ec2_metadata ami-id)"
echo "instance-id: $(get_ec2_metadata instance-id)"
echo "instance-type: $(get_ec2_metadata instance-type)"
echo "system info $(uname -a)"
- name: "[FB EMPLOYEES] Enable SSH (Click me for login details)"
uses: pytorch/test-infra/.github/actions/setup-ssh@main
continue-on-error: true
with:
github-secret: ${{ secrets.GITHUB_TOKEN }}
- name: Enable git long paths and symlinks on Windows and disable fsmonitor daemon
shell: bash
run: |
git config --global core.longpaths true
git config --global core.symlinks true
# https://git-scm.com/docs/git-fsmonitor--daemon. The daemon could lock
# the directory on Windows and prevent GHA from checking out as reported
# in https://github.com/actions/checkout/issues/1018
git config --global core.fsmonitor false
# Needed for binary builds, see: https://github.com/pytorch/pytorch/issues/73339#issuecomment-1058981560
- name: Enable long paths on Windows
shell: powershell
run: |
Set-ItemProperty -Path "HKLM:\\SYSTEM\CurrentControlSet\Control\FileSystem" -Name "LongPathsEnabled" -Value 1
# Since it's just a defensive command, the workflow should continue even the command fails. This step can be
# removed once Windows Defender is removed from the AMI
- name: Disables Windows Defender scheduled and real-time scanning for files in directories used by PyTorch
continue-on-error: true
shell: powershell
run: |
Add-MpPreference -ExclusionPath $(Get-Location).tostring(),$Env:TEMP -ErrorAction Ignore
# Let's both exclude the path and disable Windows Defender completely just to be sure
# that it doesn't interfere
Set-MpPreference -DisableRealtimeMonitoring $True -ErrorAction Ignore
- name: Checkout PyTorch
uses: actions/checkout@v4
with:
ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
submodules: recursive
path: pytorch
show-progress: false
- name: Clean PyTorch checkout
run: |
# Remove any artifacts from the previous checkouts
git clean -fxd
working-directory: pytorch
# NOTE: These environment variables are put here so that they can be applied on every job equally
# They are also here because setting them at a workflow level doesn't give us access to the
# runner.temp variable, which we need.
- name: Populate binary env
shell: bash
run: |
echo "BINARY_ENV_FILE=${RUNNER_TEMP}/env" >> "${GITHUB_ENV}"
echo "PYTORCH_FINAL_PACKAGE_DIR=${RUNNER_TEMP}/artifacts" >> "${GITHUB_ENV}"
echo "WIN_PACKAGE_WORK_DIR=${RUNNER_TEMP}"
- uses: actions/download-artifact@v4.1.7
name: Download Build Artifacts
with:
name: libtorch-cpu-shared-with-deps-debug
path: "${{ env.PYTORCH_FINAL_PACKAGE_DIR }}"
- name: Populate binary env
shell: bash
run: |
"${PYTORCH_ROOT}/.circleci/scripts/binary_populate_env.sh"
- name: Test PyTorch binary
shell: bash
run: |
"${PYTORCH_ROOT}/.circleci/scripts/binary_windows_test.sh"
- name: Wait until all sessions have drained
shell: powershell
working-directory: pytorch
if: always()
timeout-minutes: 120
run: |
.github\scripts\wait_for_ssh_to_drain.ps1
- name: Kill active ssh sessions if still around (Useful if workflow was cancelled)
shell: powershell
working-directory: pytorch
if: always()
run: |
.github\scripts\kill_active_ssh_sessions.ps1

View File

@ -1,261 +0,0 @@
# @generated DO NOT EDIT MANUALLY
# Template is at: .github/templates/windows_binary_build_workflow.yml.j2
# Generation script: .github/scripts/generate_ci_workflows.py
name: windows-binary-libtorch-release
on:
push:
branches:
- main
workflow_dispatch:
env:
# Needed for conda builds
ALPINE_IMAGE: "308535385114.dkr.ecr.us-east-1.amazonaws.com/tool/alpine"
AWS_DEFAULT_REGION: us-east-1
BUILD_ENVIRONMENT: windows-binary-libtorch-release
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
PR_NUMBER: ${{ github.event.pull_request.number }}
SHA1: ${{ github.event.pull_request.head.sha || github.sha }}
SKIP_ALL_TESTS: 1
OS: windows
concurrency:
group: windows-binary-libtorch-release-${{ github.event.pull_request.number || github.ref_name }}-${{ github.ref_type == 'branch' && github.sha }}-${{ github.event_name == 'workflow_dispatch' }}
cancel-in-progress: true
jobs:
get-label-type:
if: github.repository_owner == 'pytorch'
name: get-label-type
uses: pytorch/pytorch/.github/workflows/_runner-determinator.yml@main
with:
triggering_actor: ${{ github.triggering_actor }}
issue_owner: ${{ github.event.pull_request.user.login || github.event.issue.user.login }}
curr_branch: ${{ github.head_ref || github.ref_name }}
curr_ref_type: ${{ github.ref_type }}
libtorch-cpu-shared-with-deps-release-build:
if: ${{ github.repository_owner == 'pytorch' }}
needs: get-label-type
runs-on: "${{ needs.get-label-type.outputs.label-type }}windows.4xlarge.nonephemeral"
timeout-minutes: 360
env:
PYTORCH_ROOT: ${{ github.workspace }}/pytorch
PACKAGE_TYPE: libtorch
# TODO: This is a legacy variable that we eventually want to get rid of in
# favor of GPU_ARCH_VERSION
DESIRED_CUDA: cpu
GPU_ARCH_TYPE: cpu
SKIP_ALL_TESTS: 1
LIBTORCH_CONFIG: release
LIBTORCH_VARIANT: shared-with-deps
# This is a dummy value for libtorch to work correctly with our batch scripts
# without this value pip does not get installed for some reason
DESIRED_PYTHON: "3.10"
steps:
# NOTE: These environment variables are put here so that they can be applied on every job equally
# They are also here because setting them at a workflow level doesn't give us access to the
# runner.temp variable, which we need.
- name: Populate binary env
shell: bash
run: |
echo "BINARY_ENV_FILE=${RUNNER_TEMP}/env" >> "${GITHUB_ENV}"
echo "PYTORCH_FINAL_PACKAGE_DIR=${RUNNER_TEMP}/artifacts" >> "${GITHUB_ENV}"
echo "WIN_PACKAGE_WORK_DIR=${RUNNER_TEMP}"
- name: Display EC2 information
shell: bash
run: |
set -euo pipefail
function get_ec2_metadata() {
# Pulled from instance metadata endpoint for EC2
# see https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/instancedata-data-retrieval.html
category=$1
curl -H "X-aws-ec2-metadata-token: $(curl -s -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-ec2-metadata-token-ttl-seconds: 30")" -fsSL "http://169.254.169.254/latest/meta-data/${category}"
}
echo "ami-id: $(get_ec2_metadata ami-id)"
echo "instance-id: $(get_ec2_metadata instance-id)"
echo "instance-type: $(get_ec2_metadata instance-type)"
echo "system info $(uname -a)"
- name: "[FB EMPLOYEES] Enable SSH (Click me for login details)"
uses: pytorch/test-infra/.github/actions/setup-ssh@main
continue-on-error: true
with:
github-secret: ${{ secrets.GITHUB_TOKEN }}
- name: Enable git long paths and symlinks on Windows and disable fsmonitor daemon
shell: bash
run: |
git config --global core.longpaths true
git config --global core.symlinks true
# https://git-scm.com/docs/git-fsmonitor--daemon. The daemon could lock
# the directory on Windows and prevent GHA from checking out as reported
# in https://github.com/actions/checkout/issues/1018
git config --global core.fsmonitor false
# Needed for binary builds, see: https://github.com/pytorch/pytorch/issues/73339#issuecomment-1058981560
- name: Enable long paths on Windows
shell: powershell
run: |
Set-ItemProperty -Path "HKLM:\\SYSTEM\CurrentControlSet\Control\FileSystem" -Name "LongPathsEnabled" -Value 1
# Since it's just a defensive command, the workflow should continue even the command fails. This step can be
# removed once Windows Defender is removed from the AMI
- name: Disables Windows Defender scheduled and real-time scanning for files in directories used by PyTorch
continue-on-error: true
shell: powershell
run: |
Add-MpPreference -ExclusionPath $(Get-Location).tostring(),$Env:TEMP -ErrorAction Ignore
# Let's both exclude the path and disable Windows Defender completely just to be sure
# that it doesn't interfere
Set-MpPreference -DisableRealtimeMonitoring $True -ErrorAction Ignore
- name: Checkout PyTorch
uses: actions/checkout@v4
with:
ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
submodules: recursive
path: pytorch
show-progress: false
- name: Clean PyTorch checkout
run: |
# Remove any artifacts from the previous checkouts
git clean -fxd
working-directory: pytorch
- name: Populate binary env
shell: bash
run: |
"${PYTORCH_ROOT}/.circleci/scripts/binary_populate_env.sh"
- name: Build PyTorch binary
shell: bash
run: |
"${PYTORCH_ROOT}/.circleci/scripts/binary_windows_build.sh"
- uses: actions/upload-artifact@v4.4.0
if: always()
with:
name: libtorch-cpu-shared-with-deps-release
retention-days: 14
if-no-files-found: error
path: "${{ env.PYTORCH_FINAL_PACKAGE_DIR }}"
- name: Wait until all sessions have drained
shell: powershell
working-directory: pytorch
if: always()
timeout-minutes: 120
run: |
.github\scripts\wait_for_ssh_to_drain.ps1
- name: Kill active ssh sessions if still around (Useful if workflow was cancelled)
shell: powershell
working-directory: pytorch
if: always()
run: |
.github\scripts\kill_active_ssh_sessions.ps1
libtorch-cpu-shared-with-deps-release-test: # Testing
if: ${{ github.repository_owner == 'pytorch' }}
needs:
- libtorch-cpu-shared-with-deps-release-build
- get-label-type
runs-on: "${{ needs.get-label-type.outputs.label-type }}windows.4xlarge.nonephemeral"
timeout-minutes: 360
env:
PYTORCH_ROOT: ${{ github.workspace }}/pytorch
PACKAGE_TYPE: libtorch
# TODO: This is a legacy variable that we eventually want to get rid of in
# favor of GPU_ARCH_VERSION
DESIRED_CUDA: cpu
GPU_ARCH_TYPE: cpu
SKIP_ALL_TESTS: 1
LIBTORCH_CONFIG: release
LIBTORCH_VARIANT: shared-with-deps
# This is a dummy value for libtorch to work correctly with our batch scripts
# without this value pip does not get installed for some reason
DESIRED_PYTHON: "3.10"
steps:
- name: Display EC2 information
shell: bash
run: |
set -euo pipefail
function get_ec2_metadata() {
# Pulled from instance metadata endpoint for EC2
# see https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/instancedata-data-retrieval.html
category=$1
curl -H "X-aws-ec2-metadata-token: $(curl -s -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-ec2-metadata-token-ttl-seconds: 30")" -fsSL "http://169.254.169.254/latest/meta-data/${category}"
}
echo "ami-id: $(get_ec2_metadata ami-id)"
echo "instance-id: $(get_ec2_metadata instance-id)"
echo "instance-type: $(get_ec2_metadata instance-type)"
echo "system info $(uname -a)"
- name: "[FB EMPLOYEES] Enable SSH (Click me for login details)"
uses: pytorch/test-infra/.github/actions/setup-ssh@main
continue-on-error: true
with:
github-secret: ${{ secrets.GITHUB_TOKEN }}
- name: Enable git long paths and symlinks on Windows and disable fsmonitor daemon
shell: bash
run: |
git config --global core.longpaths true
git config --global core.symlinks true
# https://git-scm.com/docs/git-fsmonitor--daemon. The daemon could lock
# the directory on Windows and prevent GHA from checking out as reported
# in https://github.com/actions/checkout/issues/1018
git config --global core.fsmonitor false
# Needed for binary builds, see: https://github.com/pytorch/pytorch/issues/73339#issuecomment-1058981560
- name: Enable long paths on Windows
shell: powershell
run: |
Set-ItemProperty -Path "HKLM:\\SYSTEM\CurrentControlSet\Control\FileSystem" -Name "LongPathsEnabled" -Value 1
# Since it's just a defensive command, the workflow should continue even the command fails. This step can be
# removed once Windows Defender is removed from the AMI
- name: Disables Windows Defender scheduled and real-time scanning for files in directories used by PyTorch
continue-on-error: true
shell: powershell
run: |
Add-MpPreference -ExclusionPath $(Get-Location).tostring(),$Env:TEMP -ErrorAction Ignore
# Let's both exclude the path and disable Windows Defender completely just to be sure
# that it doesn't interfere
Set-MpPreference -DisableRealtimeMonitoring $True -ErrorAction Ignore
- name: Checkout PyTorch
uses: actions/checkout@v4
with:
ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
submodules: recursive
path: pytorch
show-progress: false
- name: Clean PyTorch checkout
run: |
# Remove any artifacts from the previous checkouts
git clean -fxd
working-directory: pytorch
# NOTE: These environment variables are put here so that they can be applied on every job equally
# They are also here because setting them at a workflow level doesn't give us access to the
# runner.temp variable, which we need.
- name: Populate binary env
shell: bash
run: |
echo "BINARY_ENV_FILE=${RUNNER_TEMP}/env" >> "${GITHUB_ENV}"
echo "PYTORCH_FINAL_PACKAGE_DIR=${RUNNER_TEMP}/artifacts" >> "${GITHUB_ENV}"
echo "WIN_PACKAGE_WORK_DIR=${RUNNER_TEMP}"
- uses: actions/download-artifact@v4.1.7
name: Download Build Artifacts
with:
name: libtorch-cpu-shared-with-deps-release
path: "${{ env.PYTORCH_FINAL_PACKAGE_DIR }}"
- name: Populate binary env
shell: bash
run: |
"${PYTORCH_ROOT}/.circleci/scripts/binary_populate_env.sh"
- name: Test PyTorch binary
shell: bash
run: |
"${PYTORCH_ROOT}/.circleci/scripts/binary_windows_test.sh"
- name: Wait until all sessions have drained
shell: powershell
working-directory: pytorch
if: always()
timeout-minutes: 120
run: |
.github\scripts\wait_for_ssh_to_drain.ps1
- name: Kill active ssh sessions if still around (Useful if workflow was cancelled)
shell: powershell
working-directory: pytorch
if: always()
run: |
.github\scripts\kill_active_ssh_sessions.ps1