onnxruntime/tools/ci_build/github/azure-pipelines/orttraining-linux-nightly-ortmodule-test-pipeline.yml
rui-ren b52127d22d
update acpt image for the training ci nightly (#17521)
### Description
<!-- Describe your changes. -->

The name of nightly ACPT image has been updated to
`ptebic.azurecr.io/internal/aifx/acpt/nightly-ubuntu-cuda-torch-dev`

As the previous image alias had `cu118`, `torch210dev` or `py38`, any
version update will break the training nightly pipeline



### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->

Using constant image alias to avoid pipeline failure.
2023-09-12 22:32:20 -07:00

33 lines
1.3 KiB
YAML

trigger: none
jobs:
- job: Onnxruntime_Linux_Nightly_ORTModule_tests
timeoutInMinutes: 120
pool: 'Onnxruntime-Linux-GPU-NC6sv3'
steps:
- checkout: self
clean: true
submodules: recursive
# Entry point for all ortmodule training tests
- script: |
COMMIT_ID=$(python3 -c "import onnxruntime; print(onnxruntime.get_build_info().split('git-commit-id=')[1].split(',')[0])")
cd $(Build.SourcesDirectory)
git checkout $COMMIT_ID
git branch
echo "Retrieved ONNX Runtime Commit ID: $COMMIT_ID"
docker run \
--gpus all \
--rm \
--volume $(Build.SourcesDirectory)/orttraining/orttraining/test/python:/onnxruntime_src \
--volume $(Build.SourcesDirectory)/tools/ci_build/github/linux/docker/scripts/training/ortmodule/stage1/requirements_torch_nightly:/requirements_torch_nightly \
ptebic.azurecr.io/internal/aifx/acpt/nightly-ubuntu-cuda-torch-dev \
bash -c "python3 -m pip install -r /requirements_torch_nightly/requirements.txt && python3 -m pytest -sv /onnxruntime_src/orttraining_test_ortmodule_api.py"
displayName: 'Run ORTModule Tests'
condition: succeededOrFailed()
timeoutInMinutes: 120
- template: templates/component-governance-component-detection-steps.yml
parameters:
condition: 'succeeded'