saymrwulf/onnxruntime
1
0
Fork 0
mirror of https://github.com/saymrwulf/onnxruntime.git synced 2026-05-22 22:01:08 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Move Windows GPU training job to A10 (#19041)

Browse source 
### Description
1. Update sm to 86

### Motivation and Context
We have more A10 quota then T4 and Nvidia AXX could be  partitioned
This commit is contained in:
Yi Zhang 2024-01-08 17:19:58 +00:00 committed by GitHub
parent db3c076081
commit e8ac97c8d8
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 2 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

5
tools/ci_build/github/azure-pipelines/win-gpu-ci-pipeline.yml
View file

@ -59,15 +59,14 @@ stages:
BuildConfig: 'RelWithDebInfo'
EnvSetupScript: setup_env_cuda.bat
buildArch: x64
additionalBuildFlags: --enable_pybind --enable_training --use_cuda --cuda_home="$(Agent.TempDirectory)\v11.8" --skip_onnx_tests --cmake_extra_defines CMAKE_CUDA_ARCHITECTURES=75
additionalBuildFlags: --enable_pybind --enable_training --use_cuda --cuda_home="$(Agent.TempDirectory)\v11.8" --skip_onnx_tests --cmake_extra_defines CMAKE_CUDA_ARCHITECTURES=86
msbuildPlatform: x64
isX86: false
job_name_suffix: x64_RelWithDebInfo
RunOnnxRuntimeTests: ${{ parameters.RunOnnxRuntimeTests }}
ORT_EP_NAME: CUDA
WITH_CACHE: true
# Some unit tests crash on A10 GPUs. So this job still needs to use T4.
MachinePool: onnxruntime-Win2022-GPU-T4
MachinePool: onnxruntime-Win2022-GPU-A10
isTraining: true
- stage: dml