first commit

.gitignore

@@ -0,0 +1,160 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+#poetry.lock
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#pdm.lock
+#   pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
+#   in version control.
+#   https://pdm.fming.dev/#use-with-ide
+.pdm.toml
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+#  JetBrains specific template is maintained in a separate JetBrains.gitignore that can
+#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#  and can be added to the global gitignore or merged into this file.  For a more nuclear
+#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
+#.idea/

COPYRIGHT.txt

@@ -0,0 +1 @@
+Copyright (c) 2024 MLCommons

README-test-model.md

@@ -0,0 +1,117 @@
+# CM automation for ABTF-MLPerf
+
+*Testing ABTF SSD PyTorch model via the [MLCommons CM automation meta-framework](https://github.com/mlcommons/ck).*
+
+## Install CM
+
+Follow [this online guide](https://access.cknowledge.org/playground/?action=install) to install CM for your OS.
+
+## Install virtual environment
+
+We suggest to create a virtual environment to avoid messing up your Python installation:
+
+### Linux
+
+```bash
+python3 -m venv ABTF
+. ABTF/bin/activate ; export CM_REPOS=$PWD/ABTF/CM
+```
+### Windows
+
+```bash
+python -m venv ABTF
+call ABTF\Scripts\activate.bat & set CM_REPOS=%CD%\ABTF\CM
+```
+
+## Install all CM automation recipes
+
+Pull [main MLOps automation recipes](https://access.cknowledge.org/playground/?action=scripts) from MLCommons:
+
+```bash
+cm pull repo mlcommons@ck --checkout=dev
+```
+
+Pull this CM repository with automation recipes for the MLCommons-ABTF benchmark:
+
+```bash
+cm pull repo cknowledge@cm4abtf
+```
+
+## Clean CM cache
+
+Clean CM cache if you want to start from scratch
+
+```bash
+cm rm cache -f
+```
+
+
+
+
+
+
+
+
+
+Download private test image `0000008766.png` and model `baseline_8mp.pth` to your local directory.
+
+
+Import `baseline_8mp.pth` to CM:
+```bash
+cmr "get ml-model abtf-ssd-pytorch _local.baseline_8mp.pth"
+```
+
+Get Git repo with ABTF SSD-ResNet50 PyTorch model:
+
+```bash
+cmr "get git repo _repo.https://github.com/mlcommons/abtf-ssd-pytorch" --env.CM_GIT_BRANCH=cognata-cm --extra_cache_tags=abtf,ssd,pytorch,cm-model --env.CM_GIT_CHECKOUT_PATH_ENV_NAME=CM_ABTF_SSD_PYTORCH
+```
+
+Make test prediction:
+
+```bash
+cmr "test abtf ssd-resnet50 cognata pytorch" --input=0000008766.png --output=0000008766_prediction_test.jpg --config=baseline_8MP
+```
+
+Export PyTorch model to ONNX:
+```bash
+cmr "test abtf ssd-resnet50 cognata pytorch" --input=0000008766.png --output=0000008766_prediction_test.jpg --config=baseline_8MP --export_model=baseline_8mp.onnx
+```
+
+Test exported ONNX model with LoadGen (performance):
+```bash
+cm run script "python app loadgen-generic _onnxruntime" --modelpath=baseline_8mp.onnx --samples=1 --quiet
+```
+
+
+Test different versions of PyTorch
+```bash
+cmr "install python-venv" --name=abtf2
+cmr "test abtf ssd-resnet50 cognata pytorch" --adr.python.name=abtf2 --adr.torch.version=1.13.1 --adr.torchvision.version=0.14.1 --input=0000008766.png --output=0000008766_prediction_test.jpg --config=baseline_8MP
+```
+
+## TBD
+
+### Main features
+
+* Test PyTorch model with Python LoadGen
+* Test PyTorch model with [C++ loadgen](https://github.com/mlcommons/ck/tree/master/cm-mlops/script/app-mlperf-inference-mlcommons-cpp)
+* Automate loading of Cognata dataset via CM
+* Add Cognata dataset to loadgen
+* Process PyTorch model with MLPerf inference infrastructure for SSD-ResNet50
+* Add support for MLCommons Croissant
+
+### Testing docker
+
+```bash
+cm docker script --tags=test,abtf,ssd-pytorch,_cognata --docker_cm_repo=ctuning@mlcommons-ck --env.CM_GH_TOKEN={TOKEN} --input=road.jpg --output=road_ssd.jpg
+```
+
+```bash
+cm docker script --tags=test,abtf,ssd-pytorch,_cognata --docker_cm_repo=ctuning@mlcommons-ck --docker_os=ubuntu --docker_os_version=23.04 --input=road.jpg --output=road_ssd.jpg 
+```
+TBD: pass file to CM docker: [meta](https://github.com/mlcommons/ck/blob/master/cm-mlops/script/build-mlperf-inference-server-nvidia/_cm.yaml#L197).
+
+## CM automation developers
+
+* [Grigori Fursin](https://cKnowledge.org/gfursin) (MLCommons Task Force on Automation and Reproducibility)

README.md

@@ -0,0 +1 @@
+# Collective Mind interface and automation for ABTF