[CI] [GHA] Introduce GHA Linux Conditional Compilation Pipeline #19341
Conversation
akashchi
added
category: CI
There is no log collection step in the Azure counterpart and I believe that there are no test stages that would generate artifacts of some sort to collect. |
| python3 ${{ env.OPENVINO_REPO }}/thirdparty/itt_collector/runtool/sea_runtool.py \ | ||
| --bindir ${{ env.OPENVINO_REPO }}/bin/intel64/Release -o ${{ env.BUILD_DIR }}/itt_stat ! \ | ||
| ${{ env.OPENVINO_REPO }}/bin/intel64/Release/benchmark_app -niter 1 -nireq 1 \ | ||
| -m ${{ env.MODELS_PATH }}/models/test_model/test_model_fp32.xml -d CPU |
There was a problem hiding this comment.
I recommend to use several real models instead of dummy test_modelt
We can download models from public model hubs (TF, others) and use it here.
There was a problem hiding this comment.
Is there any way we could download such model(s) w/o building and installing the Python wheels for model download and convertation? Download IRs directly from some storage?
Or do you suggest extending the workflow beyond what is in Azure with Python setup, dependencies setup, wheels building, and model downloading and convertation?
There was a problem hiding this comment.
@ilya-lavrenov , I would suggest we move the pipelines without modifications first, and improve after it's working.
…vinotoolkit#19341) * introduce linux cc * disable some triggers * check dirs * use another model * return model * rm triggers --------- Co-authored-by: Mikhail Ryzhov <[email protected]>
Details:
This PR introduces the Linux Conditional Compilation pipeline in the GitHub Actions environment. It is mirroring the Azure Linux Conditional Compilation pipeline.
Tickets: