MLPerf inference SDXL (SCC) #77
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 (SCC) | |
| on: | |
| schedule: | |
| - cron: "20 01 * * *" | |
| jobs: | |
| build_reference: | |
| if: github.repository_owner == 'gateoverflow' | |
| runs-on: [ self-hosted, linux, x64, GO-spr ] | |
| env: | |
| CM_REPOS: $HOME/GH_CM | |
| strategy: | |
| fail-fast: false | |
| matrix: | |
| python-version: [ "3.12" ] | |
| backend: [ "pytorch" ] | |
| precision: [ "float16" ] | |
| device: [ "cuda" ] | |
| steps: | |
| - name: Test MLPerf Inference reference SDXL SCC | |
| run: | | |
| if [ -f "gh_action/bin/deactivate" ]; then source gh_action/bin/deactivate; fi | |
| python3 -m venv gh_action | |
| source gh_action/bin/activate | |
| export CM_REPOS=$HOME/GH_CM | |
| pip install --upgrade cm4mlops | |
| pip install tabulate | |
| cm pull repo | |
| cm run script --tags=run-mlperf,inference,_find-performance,_r4.1-dev,_short,_scc24-base --pull_changes=yes --model=sdxl --implementation=reference --backend=${{ matrix.backend }} --category=datacenter --scenario=Offline --execution_mode=test --device=${{ matrix.device }} --precision=${{ matrix.precision }} --docker --docker_it=no --docker_cm_repo=gateoverflow@cm4mlops --docker_dt=yes --quiet --results_dir=$HOME/scc_gh_action_results --submission_dir=$HOME/scc_gh_action_submissions --precision=float16 --env.CM_MLPERF_MODEL_SDXL_DOWNLOAD_TO_HOST=yes --clean | |
| cm run script --tags=run-mlperf,inference,_r4.1-dev,_short,_scc24-base --model=sdxl --implementation=reference --backend=${{ matrix.backend }} --category=datacenter --scenario=Offline --execution_mode=test --device=${{ matrix.device }} --precision=${{ matrix.precision }} --docker --docker_it=no --docker_cm_repo=gateoverflow@cm4mlops --docker_dt=yes --quiet --results_dir=$HOME/scc_gh_action_results --submission_dir=$HOME/scc_gh_action_submissions --precision=float16 --env.CM_MLPERF_MODEL_SDXL_DOWNLOAD_TO_HOST=yes --clean | |
| cm run script --tags=generate,inference,submission --clean --run-checker --tar=yes --env.CM_TAR_OUTFILE=submission.tar.gz --division=open --category=datacenter --run_style=test --adr.submission-checker.tags=_short-run --quiet --submitter=MLCommons --submission_dir=$HOME/scc_gh_action_submissions --results_dir=$HOME/scc_gh_action_results/test_results | |
| cm run script --tags=push,github,mlperf,inference,submission --repo_url=https://github.com/gateoverflow/cm4mlperf-inference --repo_branch=mlperf-inference-results-scc24 --commit_message="Results from self hosted Github actions - NVIDIARTX4090" --quiet --submission_dir=$HOME/scc_gh_action_submissions | |
| build_nvidia: | |
| if: github.repository_owner == 'gateoverflow' | |
| runs-on: [ self-hosted, linux, x64, GO-spr] | |
| strategy: | |
| fail-fast: false | |
| matrix: | |
| python-version: [ "3.12" ] | |
| backend: [ "tensorrt" ] | |
| precision: [ "float16" ] | |
| implementation: [ "nvidia" ] | |
| steps: | |
| - name: Test MLPerf Inference NVIDIA SDXL SCC | |
| run: | | |
| if [ -f "gh_action/bin/deactivate" ]; then source gh_action/bin/deactivate; fi | |
| python3 -m venv gh_action | |
| source gh_action/bin/activate | |
| export CM_REPOS=$HOME/GH_CM | |
| pip install --upgrade cm4mlops | |
| pip install tabulate | |
| cm pull repo | |
| cm run script --tags=run-mlperf,inference,_find-performance,_r4.1-dev,_short,_scc24-base --pull_changes=yes --model=sdxl --implementation=nvidia --backend=${{ matrix.backend }} --category=datacenter --scenario=Offline --execution_mode=test --device=${{ matrix.device }} --precision=${{ matrix.precision }} --docker --docker_it=no --docker_cm_repo=gateoverflow@cm4mlops --docker_dt=yes --pull_changes --quiet --results_dir=$HOME/scc_gh_action_results --submission_dir=$HOME/scc_gh_action_submissions --env.CM_MLPERF_MODEL_SDXL_DOWNLOAD_TO_HOST=yes --hw_name=go-spr --custom_system_nvidia=yes --clean | |
| cm run script --tags=run-mlperf,inference,_r4.1-dev,_short,_scc24-base --model=sdxl --implementation=nvidia --backend=${{ matrix.backend }} --category=datacenter --scenario=Offline --execution_mode=test --device=${{ matrix.device }} --precision=${{ matrix.precision }} --docker --docker_it=no --docker_cm_repo=gateoverflow@cm4mlops --docker_dt=yes --quiet --results_dir=$HOME/scc_gh_action_results --submission_dir=$HOME/scc_gh_action_submissions --precision=float16 --env.CM_MLPERF_MODEL_SDXL_DOWNLOAD_TO_HOST=yes --clean | |
| cm run script --tags=generate,inference,submission --clean --run-checker --tar=yes --env.CM_TAR_OUTFILE=submission.tar.gz --division=open --category=datacenter --run_style=test --adr.submission-checker.tags=_short-run --quiet --submitter=MLCommons --submission_dir=$HOME/scc_gh_action_submissions --results_dir=$HOME/scc_gh_action_results/test_results | |
| cm run script --tags=push,github,mlperf,inference,submission --repo_url=https://github.com/gateoverflow/cm4mlperf-inference --repo_branch=mlperf-inference-results-scc24 --commit_message="Results from self hosted Github actions - NVIDIARTX4090" --quiet --submission_dir=$HOME/scc_gh_action_submissions |