added workflow for nvidia #1
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| name: MLPerf inference SDXL | ||
| on: | ||
| schedule: | ||
| - cron: "1 1 * * */3" | ||
| jobs: | ||
| build_reference: | ||
| if: github.repository_owner == 'gateoverflow' | ||
| runs-on: [ self-hosted, linux, x64 ] | ||
| strategy: | ||
| fail-fast: false | ||
| matrix: | ||
| python-version: [ "3.12" ] | ||
| backend: [ "pytorch" ] | ||
| precision: [ "float16" ] | ||
| steps: | ||
| - name: Install dependencies | ||
| run: | | ||
| source gh_action/bin/deactivate || python3 -m venv gh_action | ||
| source gh_action/bin/activate | ||
| export CM_REPOS=$HOME/GH_CM | ||
| cm pull repo --url=${{ github.event.pull_request.head.repo.html_url }} --checkout=${{ github.event.pull_request.head.ref }} | ||
| - name: Test MLPerf Inference SDXL | ||
| run: | | ||
| cm run script --tags=run-mlperf,inference,_submission,_short --submitter="MLCommons" --docker --model=sdxl --backend=${{ matrix.backend }} --device=cuda --scenario=Offline --test_query_count=1 --precision=${{ matrix.precision }} --target_qps=1 --quiet --docker_it=no --docker_cm_repo=gateoverflow@cm4mlops --adr.compiler.tags=gcc --beam_size=1 --hw_name=gh_action --docker_dt=yes --results_dir=$HOME/gh_action_results --submission_dir=$HOME/gh_action_submissions --clean | ||
| build_nvidia: | ||
| if: github.repository_owner == 'gateoverflow' | ||
| runs-on: [ self-hosted, linux, x64 ] | ||
| strategy: | ||
| fail-fast: false | ||
| matrix: | ||
| python-version: [ "3.12" ] | ||
| backend: [ "tensorrt" ] | ||
| precision: [ "float16" ] | ||
| implementation: [ "nvidia" ] | ||
| steps: | ||
| - name: Install dependencies | ||
| run: | | ||
| source gh_action/bin/deactivate || python3 -m venv gh_action | ||
| source gh_action/bin/activate | ||
| export CM_REPOS=$HOME/GH_CM | ||
| cm pull repo --url=${{ github.event.pull_request.head.repo.html_url }} --checkout=${{ github.event.pull_request.head.ref }} | ||
| - name: Test MLPerf Inference SDXL | ||
| run: | | ||
| cm run script --tags=run-mlperf,inference,_submission,_short --submitter="MLCommons" --docker --model=sdxl --implementation=${{ matrix.implementation }} --backend=${{ matrix.backend }} --device=cuda --scenario=Offline --test_query_count=1 --precision=${{ matrix.precision }} --target_qps=1 --quiet --docker_it=no --docker_cm_repo=gateoverflow@cm4mlops --adr.compiler.tags=gcc --beam_size=1 --hw_name=gh_action --docker_dt=yes --results_dir=$HOME/gh_action_results --submission_dir=$HOME/gh_action_submissions --clean | ||
