OSSForks/pytorch
1
0
Fork 0
mirror of https://github.com/zebrajr/pytorch.git synced 2025-12-06 12:20:52 +01:00
Code Issues Actions 38 Packages Projects Releases Wiki Activity

Move cuda 12.4 jobs to periodic for both pull and inductor (#127825)

Browse Source 
Moves 12.4 sm86/a10g jobs in pull to trunk
Moves 12.4 cuda non sm86 jobs to periodic
Moves 12.4 jobs in inductor to inductor-periodic, except inductor_timm which seems to give important signal

There has been a lot of queueing for cuda runners due to the addition of jobs for cuda 12.4, so move those jobs to other workflows that are run less often
Co-authored-by: Andrey Talman <atalman@fb.com>
Pull Request resolved: https://github.com/pytorch/pytorch/pull/127825
Approved by: https://github.com/ZainRizvi, https://github.com/nWEIdia, https://github.com/atalman, https://github.com/malfet
This commit is contained in:
Catherine Lee 2024-06-05 21:01:36 +00:00 committed by PyTorch MergeBot
parent 8184cd85fc
commit 01694eaa56
5 changed files with 128 additions and 121 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

90
.github/workflows/inductor-periodic.yml vendored
View File

@ -56,3 +56,93 @@ jobs:
test-matrix: ${{ needs.linux-focal-cuda12_1-py3_10-gcc9-periodic-dynamo-benchmarks-build.outputs.test-matrix }}
secrets:
HUGGING_FACE_HUB_TOKEN: ${{ secrets.HUGGING_FACE_HUB_TOKEN }}
linux-focal-cuda12_4-py3_10-gcc9-inductor-build:
# Should be synced with the one in inductor.yml, but this doesn't run inductor_timm
name: cuda12.4-py3.10-gcc9-sm86
uses: ./.github/workflows/_linux-build.yml
with:
sync-tag: linux-focal-cuda12_4-py3_10-gcc9-inductor-build
build-environment: linux-focal-cuda12.4-py3.10-gcc9-sm86
docker-image-name: pytorch-linux-focal-cuda12.4-cudnn8-py3-gcc9-inductor-benchmarks
cuda-arch-list: '8.6'
test-matrix: |
{ include: [
{ config: "inductor", shard: 1, num_shards: 1, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "inductor_distributed", shard: 1, num_shards: 1, runner: "linux.g5.12xlarge.nvidia.gpu" },
{ config: "inductor_huggingface", shard: 1, num_shards: 1, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "inductor_torchbench", shard: 1, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "inductor_torchbench", shard: 2, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "dynamic_inductor_huggingface", shard: 1, num_shards: 1, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "dynamic_inductor_timm", shard: 1, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "dynamic_inductor_timm", shard: 2, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "dynamic_inductor_torchbench", shard: 1, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "dynamic_inductor_torchbench", shard: 2, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "aot_inductor_huggingface", shard: 1, num_shards: 1, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "aot_inductor_timm", shard: 1, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "aot_inductor_timm", shard: 2, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "aot_inductor_torchbench", shard: 1, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "aot_inductor_torchbench", shard: 2, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "inductor_cpp_wrapper_abi_compatible", shard: 1, num_shards: 1, runner: "linux.g5.4xlarge.nvidia.gpu" },
]}
secrets:
HUGGING_FACE_HUB_TOKEN: ${{ secrets.HUGGING_FACE_HUB_TOKEN }}
linux-focal-cuda12_4-py3_10-gcc9-inductor-test:
name: cuda12.4-py3.10-gcc9-sm86
uses: ./.github/workflows/_linux-test.yml
needs: linux-focal-cuda12_4-py3_10-gcc9-inductor-build
with:
sync-tag: linux-focal-cuda12_4-py3_10-gcc9-inductor-test
build-environment: linux-focal-cuda12.4-py3.10-gcc9-sm86
docker-image: ${{ needs.linux-focal-cuda12_4-py3_10-gcc9-inductor-build.outputs.docker-image }}
test-matrix: ${{ needs.linux-focal-cuda12_4-py3_10-gcc9-inductor-build.outputs.test-matrix }}
secrets:
HUGGING_FACE_HUB_TOKEN: ${{ secrets.HUGGING_FACE_HUB_TOKEN }}
linux-focal-cuda12_4-py3_10-gcc9-inductor-build-gcp:
name: cuda12.4-py3.10-gcc9-sm80
uses: ./.github/workflows/_linux-build.yml
with:
build-environment: linux-focal-cuda12.4-py3.10-gcc9-sm80
docker-image-name: pytorch-linux-focal-cuda12.4-cudnn8-py3-gcc9-inductor-benchmarks
cuda-arch-list: '8.0'
test-matrix: |
{ include: [
{ config: "inductor_torchbench_smoketest_perf", shard: 1, num_shards: 1, runner: "linux.gcp.a100" },
]}
secrets:
HUGGING_FACE_HUB_TOKEN: ${{ secrets.HUGGING_FACE_HUB_TOKEN }}
linux-focal-cuda12_4-py3_10-gcc9-inductor-test-gcp:
name: cuda12.4-py3.10-gcc9-sm80
uses: ./.github/workflows/_linux-test.yml
needs: linux-focal-cuda12_4-py3_10-gcc9-inductor-build-gcp
with:
build-environment: linux-focal-cuda12.4-py3.10-gcc9-sm80
docker-image: ${{ needs.linux-focal-cuda12_4-py3_10-gcc9-inductor-build-gcp.outputs.docker-image }}
test-matrix: ${{ needs.linux-focal-cuda12_4-py3_10-gcc9-inductor-build-gcp.outputs.test-matrix }}
use-gha: anything-non-empty-to-use-gha
secrets:
HUGGING_FACE_HUB_TOKEN: ${{ secrets.HUGGING_FACE_HUB_TOKEN }}
linux-focal-cuda12_4-py3_12-gcc9-inductor-build:
name: cuda12.4-py3.12-gcc9-sm86
uses: ./.github/workflows/_linux-build.yml
with:
build-environment: linux-focal-cuda12.4-py3.12-gcc9-sm86
docker-image-name: pytorch-linux-focal-cuda12.4-cudnn8-py3.12-gcc9-inductor-benchmarks
cuda-arch-list: '8.6'
test-matrix: |
{ include: [
{ config: "inductor", shard: 1, num_shards: 1, runner: "linux.g5.4xlarge.nvidia.gpu" },
]}
linux-focal-cuda12_4-py3_12-gcc9-inductor-test:
name: cuda12.4-py3.12-gcc9-sm86
uses: ./.github/workflows/_linux-test.yml
needs: linux-focal-cuda12_4-py3_12-gcc9-inductor-build
with:
build-environment: linux-focal-cuda12.4-py3.12-gcc9-sm86
docker-image: ${{ needs.linux-focal-cuda12_4-py3_12-gcc9-inductor-build.outputs.docker-image }}
test-matrix: ${{ needs.linux-focal-cuda12_4-py3_12-gcc9-inductor-build.outputs.test-matrix }}

66
.github/workflows/inductor.yml vendored
View File

@ -129,32 +129,18 @@ jobs:
test-matrix: ${{ needs.linux-focal-cuda12_1-py3_12-gcc9-inductor-build.outputs.test-matrix }}
linux-focal-cuda12_4-py3_10-gcc9-inductor-build:
# Should be synced with the one in inductor-periodic.yml but this only runs inductor_timm
name: cuda12.4-py3.10-gcc9-sm86
uses: ./.github/workflows/_linux-build.yml
with:
sync-tag: linux-focal-cuda12_4-py3_10-gcc9-inductor-build
build-environment: linux-focal-cuda12.4-py3.10-gcc9-sm86
docker-image-name: pytorch-linux-focal-cuda12.4-cudnn8-py3-gcc9-inductor-benchmarks
cuda-arch-list: '8.6'
test-matrix: |
{ include: [
{ config: "inductor", shard: 1, num_shards: 1, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "inductor_distributed", shard: 1, num_shards: 1, runner: "linux.g5.12xlarge.nvidia.gpu" },
{ config: "inductor_huggingface", shard: 1, num_shards: 1, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "inductor_timm", shard: 1, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "inductor_timm", shard: 2, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "inductor_torchbench", shard: 1, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "inductor_torchbench", shard: 2, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "dynamic_inductor_huggingface", shard: 1, num_shards: 1, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "dynamic_inductor_timm", shard: 1, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "dynamic_inductor_timm", shard: 2, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "dynamic_inductor_torchbench", shard: 1, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "dynamic_inductor_torchbench", shard: 2, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "aot_inductor_huggingface", shard: 1, num_shards: 1, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "aot_inductor_timm", shard: 1, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "aot_inductor_timm", shard: 2, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "aot_inductor_torchbench", shard: 1, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "aot_inductor_torchbench", shard: 2, num_shards: 2, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "inductor_cpp_wrapper_abi_compatible", shard: 1, num_shards: 1, runner: "linux.g5.4xlarge.nvidia.gpu" },
]}
secrets:
HUGGING_FACE_HUB_TOKEN: ${{ secrets.HUGGING_FACE_HUB_TOKEN }}
@ -164,59 +150,13 @@ jobs:
uses: ./.github/workflows/_linux-test.yml
needs: linux-focal-cuda12_4-py3_10-gcc9-inductor-build
with:
sync-tag: linux-focal-cuda12_4-py3_10-gcc9-inductor-test
build-environment: linux-focal-cuda12.4-py3.10-gcc9-sm86
docker-image: ${{ needs.linux-focal-cuda12_4-py3_10-gcc9-inductor-build.outputs.docker-image }}
test-matrix: ${{ needs.linux-focal-cuda12_4-py3_10-gcc9-inductor-build.outputs.test-matrix }}
secrets:
HUGGING_FACE_HUB_TOKEN: ${{ secrets.HUGGING_FACE_HUB_TOKEN }}
linux-focal-cuda12_4-py3_10-gcc9-inductor-build-gcp:
name: cuda12.4-py3.10-gcc9-sm80
uses: ./.github/workflows/_linux-build.yml
with:
build-environment: linux-focal-cuda12.4-py3.10-gcc9-sm80
docker-image-name: pytorch-linux-focal-cuda12.4-cudnn8-py3-gcc9-inductor-benchmarks
cuda-arch-list: '8.0'
test-matrix: |
{ include: [
{ config: "inductor_torchbench_smoketest_perf", shard: 1, num_shards: 1, runner: "linux.gcp.a100" },
]}
secrets:
HUGGING_FACE_HUB_TOKEN: ${{ secrets.HUGGING_FACE_HUB_TOKEN }}
linux-focal-cuda12_4-py3_12-gcc9-inductor-build:
name: cuda12.4-py3.12-gcc9-sm86
uses: ./.github/workflows/_linux-build.yml
with:
build-environment: linux-focal-cuda12.4-py3.12-gcc9-sm86
docker-image-name: pytorch-linux-focal-cuda12.4-cudnn8-py3.12-gcc9-inductor-benchmarks
cuda-arch-list: '8.6'
test-matrix: |
{ include: [
{ config: "inductor", shard: 1, num_shards: 1, runner: "linux.g5.4xlarge.nvidia.gpu" },
]}
linux-focal-cuda12_4-py3_10-gcc9-inductor-test-gcp:
name: cuda12.4-py3.10-gcc9-sm80
uses: ./.github/workflows/_linux-test.yml
needs: linux-focal-cuda12_4-py3_10-gcc9-inductor-build-gcp
with:
build-environment: linux-focal-cuda12.4-py3.10-gcc9-sm80
docker-image: ${{ needs.linux-focal-cuda12_4-py3_10-gcc9-inductor-build-gcp.outputs.docker-image }}
test-matrix: ${{ needs.linux-focal-cuda12_4-py3_10-gcc9-inductor-build-gcp.outputs.test-matrix }}
use-gha: anything-non-empty-to-use-gha
secrets:
HUGGING_FACE_HUB_TOKEN: ${{ secrets.HUGGING_FACE_HUB_TOKEN }}
linux-focal-cuda12_4-py3_12-gcc9-inductor-test:
name: cuda12.4-py3.12-gcc9-sm86
uses: ./.github/workflows/_linux-test.yml
needs: linux-focal-cuda12_4-py3_12-gcc9-inductor-build
with:
build-environment: linux-focal-cuda12.4-py3.12-gcc9-sm86
docker-image: ${{ needs.linux-focal-cuda12_4-py3_12-gcc9-inductor-build.outputs.docker-image }}
test-matrix: ${{ needs.linux-focal-cuda12_4-py3_12-gcc9-inductor-build.outputs.test-matrix }}
linux-jammy-cpu-py3_8-gcc11-inductor-build:
name: linux-jammy-cpu-py3.8-gcc11-inductor
uses: ./.github/workflows/_linux-build.yml

11
.github/workflows/periodic.yml vendored
View File

@ -37,7 +37,6 @@ jobs:
permissions:
id-token: write
contents: read
linux-focal-cuda12_1-py3_10-gcc9-build:
name: linux-focal-cuda12.1-py3.10-gcc9
uses: ./.github/workflows/_linux-build.yml
@ -50,7 +49,6 @@ jobs:
{ config: "nogpu_NO_AVX2", shard: 1, num_shards: 1, runner: "linux.2xlarge" },
{ config: "jit_legacy", shard: 1, num_shards: 1, runner: "linux.4xlarge.nvidia.gpu" },
]}
linux-focal-cuda12_1-py3_10-gcc9-test:
name: linux-focal-cuda12.1-py3.10-gcc9
uses: ./.github/workflows/_linux-test.yml
@ -64,12 +62,18 @@ jobs:
linux-focal-cuda12_4-py3_10-gcc9-build:
name: linux-focal-cuda12.4-py3.10-gcc9
uses: ./.github/workflows/_linux-build.yml
uses: ./.github/workflows/_linux-build-label.yml
with:
build-environment: linux-focal-cuda12.4-py3.10-gcc9
docker-image-name: pytorch-linux-focal-cuda12.4-cudnn8-py3-gcc9
test-matrix: |
{ include: [
{ config: "default", shard: 1, num_shards: 5, runner: "linux.4xlarge.nvidia.gpu" },
{ config: "default", shard: 2, num_shards: 5, runner: "linux.4xlarge.nvidia.gpu" },
{ config: "default", shard: 3, num_shards: 5, runner: "linux.4xlarge.nvidia.gpu" },
{ config: "default", shard: 4, num_shards: 5, runner: "linux.4xlarge.nvidia.gpu" },
{ config: "default", shard: 5, num_shards: 5, runner: "linux.4xlarge.nvidia.gpu" },
{ config: "deploy", shard: 1, num_shards: 1, runner: "linux.4xlarge.nvidia.gpu" },
{ config: "nogpu_AVX512", shard: 1, num_shards: 1, runner: "linux.2xlarge" },
{ config: "nogpu_NO_AVX2", shard: 1, num_shards: 1, runner: "linux.2xlarge" },
{ config: "jit_legacy", shard: 1, num_shards: 1, runner: "linux.4xlarge.nvidia.gpu" },
@ -82,6 +86,7 @@ jobs:
- linux-focal-cuda12_4-py3_10-gcc9-build
- target-determination
with:
timeout-minutes: 360
build-environment: linux-focal-cuda12.4-py3.10-gcc9
docker-image: ${{ needs.linux-focal-cuda12_4-py3_10-gcc9-build.outputs.docker-image }}
test-matrix: ${{ needs.linux-focal-cuda12_4-py3_10-gcc9-build.outputs.test-matrix }}

55
.github/workflows/pull.yml vendored
View File

@ -285,34 +285,6 @@ jobs:
docker-image: ${{ needs.linux-focal-cuda12_1-py3_10-gcc9-build.outputs.docker-image }}
test-matrix: ${{ needs.linux-focal-cuda12_1-py3_10-gcc9-build.outputs.test-matrix }}
linux-focal-cuda12_4-py3_10-gcc9-build:
name: linux-focal-cuda12.4-py3.10-gcc9
uses: ./.github/workflows/_linux-build-label.yml
with:
build-environment: linux-focal-cuda12.4-py3.10-gcc9
docker-image-name: pytorch-linux-focal-cuda12.4-cudnn8-py3-gcc9
test-matrix: |
{ include: [
{ config: "default", shard: 1, num_shards: 5, runner: "linux.4xlarge.nvidia.gpu" },
{ config: "default", shard: 2, num_shards: 5, runner: "linux.4xlarge.nvidia.gpu" },
{ config: "default", shard: 3, num_shards: 5, runner: "linux.4xlarge.nvidia.gpu" },
{ config: "default", shard: 4, num_shards: 5, runner: "linux.4xlarge.nvidia.gpu" },
{ config: "default", shard: 5, num_shards: 5, runner: "linux.4xlarge.nvidia.gpu" },
{ config: "deploy", shard: 1, num_shards: 1, runner: "linux.4xlarge.nvidia.gpu" },
]}
linux-focal-cuda12_4-py3_10-gcc9-test:
name: linux-focal-cuda12.4-py3.10-gcc9
uses: ./.github/workflows/_linux-test.yml
needs:
- linux-focal-cuda12_4-py3_10-gcc9-build
- target-determination
with:
timeout-minutes: 360
build-environment: linux-focal-cuda12.4-py3.10-gcc9
docker-image: ${{ needs.linux-focal-cuda12_4-py3_10-gcc9-build.outputs.docker-image }}
test-matrix: ${{ needs.linux-focal-cuda12_4-py3_10-gcc9-build.outputs.test-matrix }}
linux-jammy-py3-clang12-mobile-build:
name: linux-jammy-py3-clang12-mobile-build
uses: ./.github/workflows/_linux-build-label.yml
@ -497,33 +469,6 @@ jobs:
docker-image: ${{ needs.linux-focal-cuda12_1-py3_10-gcc9-sm86-build.outputs.docker-image }}
test-matrix: ${{ needs.linux-focal-cuda12_1-py3_10-gcc9-sm86-build.outputs.test-matrix }}
linux-focal-cuda12_4-py3_10-gcc9-sm86-build:
name: linux-focal-cuda12.4-py3.10-gcc9-sm86
uses: ./.github/workflows/_linux-build-label.yml
with:
build-environment: linux-focal-cuda12.4-py3.10-gcc9-sm86
docker-image-name: pytorch-linux-focal-cuda12.4-cudnn8-py3-gcc9
cuda-arch-list: 8.6
test-matrix: |
{ include: [
{ config: "default", shard: 1, num_shards: 5, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "default", shard: 2, num_shards: 5, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "default", shard: 3, num_shards: 5, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "default", shard: 4, num_shards: 5, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "default", shard: 5, num_shards: 5, runner: "linux.g5.4xlarge.nvidia.gpu" },
]}
linux-focal-cuda12_4-py3_10-gcc9-sm86-test:
name: linux-focal-cuda12.4-py3.10-gcc9-sm86
uses: ./.github/workflows/_linux-test.yml
needs:
- linux-focal-cuda12_4-py3_10-gcc9-sm86-build
- target-determination
with:
build-environment: linux-focal-cuda12.4-py3.10-gcc9-sm86
docker-image: ${{ needs.linux-focal-cuda12_4-py3_10-gcc9-sm86-build.outputs.docker-image }}
test-matrix: ${{ needs.linux-focal-cuda12_4-py3_10-gcc9-sm86-build.outputs.test-matrix }}
linux-jammy-py3-clang12-executorch-build:
name: linux-jammy-py3-clang12-executorch
uses: ./.github/workflows/_linux-build-label.yml

27
.github/workflows/trunk.yml vendored
View File

@ -34,6 +34,33 @@ jobs:
id-token: write
contents: read
linux-focal-cuda12_4-py3_10-gcc9-sm86-build:
name: linux-focal-cuda12.4-py3.10-gcc9-sm86
uses: ./.github/workflows/_linux-build-label.yml
with:
build-environment: linux-focal-cuda12.4-py3.10-gcc9-sm86
docker-image-name: pytorch-linux-focal-cuda12.4-cudnn8-py3-gcc9
cuda-arch-list: 8.6
test-matrix: |
{ include: [
{ config: "default", shard: 1, num_shards: 5, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "default", shard: 2, num_shards: 5, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "default", shard: 3, num_shards: 5, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "default", shard: 4, num_shards: 5, runner: "linux.g5.4xlarge.nvidia.gpu" },
{ config: "default", shard: 5, num_shards: 5, runner: "linux.g5.4xlarge.nvidia.gpu" },
]}
linux-focal-cuda12_4-py3_10-gcc9-sm86-test:
name: linux-focal-cuda12.4-py3.10-gcc9-sm86
uses: ./.github/workflows/_linux-test.yml
needs:
- linux-focal-cuda12_4-py3_10-gcc9-sm86-build
- target-determination
with:
build-environment: linux-focal-cuda12.4-py3.10-gcc9-sm86
docker-image: ${{ needs.linux-focal-cuda12_4-py3_10-gcc9-sm86-build.outputs.docker-image }}
test-matrix: ${{ needs.linux-focal-cuda12_4-py3_10-gcc9-sm86-build.outputs.test-matrix }}
libtorch-linux-focal-cuda12_1-py3_7-gcc9-debug-build:
name: libtorch-linux-focal-cuda12.1-py3.7-gcc9-debug
uses: ./.github/workflows/_linux-build.yml