Add Workflow Docs

.github/dockerhub-readmes.json

@@ -39,6 +39,10 @@
         {
             "fname": "./preset/README.md",
             "repo-name": "intel/inference-optimization"
+        },
+        {
+            "fname": "./workflows/README.md",
+            "repo-name": "intel/ai-workflows"
         }
     ]
 }

README.md

@@ -2,9 +2,9 @@
 
 [![OpenSSF Best Practices](https://www.bestpractices.dev/projects/8270/badge)](https://www.bestpractices.dev/projects/8270)
 [![OpenSSF Scorecard](https://api.securityscorecards.dev/projects/github.com/intel/ai-containers/badge)](https://securityscorecards.dev/viewer/?uri=github.com/intel/ai-containers)
-[![FOSSA Status](https://app.fossa.com/api/projects/git%2Bgithub.com%2Fintel%2Fai-containers.svg?type=shield&issueType=license)](https://app.fossa.com/projects/git%2Bgithub.com%2Fintel%2Fai-containers?ref=badge_shield&issueType=license)
-[![CodeQL](https://github.com/intel/ai-containers/actions/workflows/github-code-scanning/codeql/badge.svg)](https://github.com/intel/ai-containers/actions/workflows/github-code-scanning/codeql)
 [![pre-commit.ci status](https://results.pre-commit.ci/badge/github/intel/ai-containers/main.svg)](https://results.pre-commit.ci/latest/github/intel/ai-containers/main)
+[![CodeQL](https://github.com/intel/ai-containers/actions/workflows/github-code-scanning/codeql/badge.svg)](https://github.com/intel/ai-containers/actions/workflows/github-code-scanning/codeql)
+[![Lint](https://github.com/intel/ai-containers/actions/workflows/lint.yaml/badge.svg)](https://github.com/intel/ai-containers/actions/workflows/lint.yaml)
 [![Test Runner CI](https://github.com/intel/ai-containers/actions/workflows/test-runner-ci.yaml/badge.svg)](https://github.com/intel/ai-containers/actions/workflows/test-runner-ci.yaml)
 [![Helm Chart CI](https://github.com/intel/ai-containers/actions/workflows/chart-ci.yaml/badge.svg)](https://github.com/intel/ai-containers/actions/workflows/chart-ci.yaml)
 [![Weekly Tests](https://github.com/intel/ai-containers/actions/workflows/weekly-test.yaml/badge.svg)](https://github.com/intel/ai-containers/actions/workflows/weekly-test.yaml)
@@ -17,14 +17,18 @@ Define your project's registry and repository each time you use the project:
 
 ```bash
 # REGISTRY/REPO:TAG
-export CACHE_REGISTRY=<cache_registry_name>
 export REGISTRY=<registry_name>
 export REPO=<repo_name>
 ```
 
+The maintainers of Intel® AI Containers use [harbor](https://github.com/goharbor/harbor) to store containers.
+
 > [!NOTE]
 > `REGISTRY` and `REPO` are used to authenticate with the private registry necessary to push completed container layers and saved them for testing and publication. For example: `REGISTRY=intel && REPO=intel-extension-for-pytorch` would become `intel/intel-extension-for-pytorch` as the name of the container image, followed by the tag generated from the service found in that project's compose file.
 
+> [!WARNING]
+> You can optionally skip this step and use some placeholder values, however some container groups depend on other images and will pull from a registry that you have not defined and result in an error.
+
 ### Set Up Docker Engine
 
 You'll need to install Docker Engine on your development system. Note that while **Docker Engine** is free to use, **Docker Desktop** may require you to purchase a license.  See the [Docker Engine Server installation instructions](https://docs.docker.com/engine/install/#server) for details.
@@ -61,6 +65,9 @@ cd pytorch
 PACKAGE_OPTION=idp docker compose up --build ipex-base
 ```
 
+> [!NOTE]
+> If you didn't specify `REGISTRY` or `REPO`, you also need to add the `idp` service to the list to build the dependent python image.
+
 ## Test Containers
 
 To test the containers, use the [Test Runner Framework](https://github.com/intel/ai-containers/tree/main/test-runner):

docs/README.md

@@ -24,6 +24,9 @@ mkdocs serve --no-livereload
 
 The documentation will be available at [http://localhost:8000](http://localhost:8000).
 
+> [!CAUTION]
+> Docker compose `v2.26.1` is the minimum required version.
+
 ## Labelling Containers
 
 To customize the tables in the [Support Matrix](./matrix.md), you can add labels to the services found in each container group's `docker-compose.yaml` file. The command `docker compose config` is run to get all of the metadata from the container group. Labels are used to specify the public metadata for each container, and then the tables are generated based on the `.actions.json` file found in the same directory.

docs/matrix.md

@@ -1,13 +1,13 @@
 # Support Matrix
 
 === "Framework Containers"
-    === "Python"
+    === "[Python](https://hub.docker.com/r/intel/python)"
         {{ read_csv('assets/python.csv') }}
-    === "Classical ML"
+    === "[Classical ML](https://hub.docker.com/r/intel/intel-optimized-ml)"
         {{ read_csv('assets/classical-ml.csv') }}
-    === "PyTorch"
+    === "[PyTorch](https://hub.docker.com/r/intel/intel-optimized-pytorch)"
         {{ read_csv('assets/pytorch.csv') }}
-    === "TensorFlow"
+    === "[TensorFlow](https://hub.docker.com/r/intel/intel-optimized-tensorflow)"
         {{ read_csv('assets/tensorflow.csv') }}
 
 === "Model Containers"
@@ -22,7 +22,7 @@
         === "Max"
             {{ read_csv('assets/max-pvc.csv') }}
 
-=== "Preset Containers"
+=== "[Preset Containers](https://github.com/intel/ai-containers/blob/main/preset/README.md)"
     === "Data Analytics"
         {{ read_csv('assets/data_analytics.csv') }}
     === "Classical ML"
@@ -32,10 +32,8 @@
     === "Inference Optimization"
         {{ read_csv('assets/inference_optimization.csv') }}
 
-=== "Other"
-    === "Serving"
+=== "[Workflows](https://hub.docker.com/r/intel/ai-workflows)"
+    === "[TorchServe](https://github.com/intel/ai-containers/tree/main/workflows/charts/torchserve)"
         {{ read_csv('assets/serving.csv') }}
-    === "Transformers"
-        {{ read_csv('assets/transformers.csv') }}
-    === "GenAI"
+    === "[Huggingface LLM](https://github.com/intel/ai-containers/tree/main/workflows/charts/huggingface-llm)"
         {{ read_csv('assets/genai.csv') }}

docs/scripts/hook.py

@@ -40,8 +40,7 @@ def create_support_matrix():
     # compose_to_csv("preset/classical-ml", "classical_ml")
     # compose_to_csv("preset/deep-learning", "deep_learning")
     # compose_to_csv("preset/inference-optimization", "inference_optimization")
-    # get_repo(transformers)
-    # get_repo(genai)
+    compose_to_csv("workflows/charts/huggingface-llm", "genai")
 
 
 def on_pre_build(*args, **kwargs):

docs/scripts/matrix.py

@@ -174,7 +174,8 @@ def make_table(setting: str = None, compose_metadata: dict = None):
             metadata = extract_labels(setting)
             df = pd.concat([df, make_table(setting, metadata)], axis=1)
     else:
-        df = make_table()
+        metadata = docker.compose.config(return_json=True)
+        df = make_table("Name=Version", metadata)
 
     os.chdir(root)
     df.loc[:, ~df.columns.duplicated()].to_csv(f"docs/assets/{path}.csv", index=False)

docs/scripts/readmes.py

@@ -20,8 +20,8 @@
     "preset/README.md",
     "python/README.md",
     "pytorch/README.md",
-    "pytorch/serving/README.md",
     "tensorflow/README.md",
+    "workflows/README.md",
 ]
 
 

mkdocs.yml

@@ -13,7 +13,6 @@
 # limitations under the License.
 
 copyright: Copyright © 2024 Intel Corporation
-edit_uri: edit/main/docs/
 extra:
   generator: false
 extra_javascript:
@@ -37,15 +36,11 @@ nav:
   - Python Base: 'python/README.md'
   - PyTorch Base: 'pytorch/README.md'
   - TensorFlow Base: 'tensorflow/README.md'
-  - TorchServe: 'pytorch/serving/README.md'
+  - Workflows: 'workflows/README.md'
 - Support Matrix: 'matrix.md'
 - Roadmap: 'roadmap.md'
 plugins:
 - callouts
-- git-authors
-- git-revision-date-localized:
-    enable_creation_date: true
-    type: date
 # - optimize
 - search
 - table-reader:
@@ -73,5 +68,4 @@ theme:
   - navigation.prune
   - toc.follow
   - toc.integrate
-  - content.action.edit
   name: material

test-runner/README.md

@@ -277,3 +277,16 @@ See an [Example](../.github/workflows/test-runner-ci.yaml#L94) Implementation of
 
 > [!TIP]
 > When writing Tests for use with a CI platform like GitHub Actions, write your tests in such a way that they would be executed from the root directory of your repository.
+
+## Testing
+
+For testing [tests.yaml](tests.yaml) file, some variables need to be set
+
+```bash
+export CACHE_REGISTRY=<harbor cache_registry_name>
+export REGISTRY=<harbor registry>
+export REPO=<harbor project/repo>
+# optional
+export PERF_REPO=<internal perf repo>
+python test-runner/test_runner.py -f test-runner/tests.yaml -a test-runner/.actions.json
+```

workflows/README.md

@@ -1,10 +1,36 @@
-# Workflows
+# Intel® AI Workflows
 
-This directory contains workflows demonstrating showing how the Intel Optimized base containers can be used for
-different use cases:
+Demonstrating showing how the [Intel® AI Containers] can be used for different use cases:
 
 ## PyTorch Workflows
 
 | Base Container | Device Type | Example | Description |
 |----------------|-------------|---------|-------------|
-| `intel/intel-optimized-pytorch:2.3.0-pip-multinode` | CPU | [Distributed LLM Fine Tuning with Kubernetes](charts/huggingface-llm) | Demonstrates using Hugging Face Transformers with Intel® Xeon® Scalable Processors to fine tune LLMs with multiple nodes from a Kubernetes cluster. The example includes a LLM fine tuning script, Dockerfile, and Helm chart. |
+| `intel/intel-optimized-pytorch:2.3.0-pip-multinode` | CPU | [Distributed LLM Fine Tuning with Kubernetes] | Demonstrates using Hugging Face Transformers with Intel® Xeon® Scalable Processors to fine tune LLMs with multiple nodes from a Kubernetes cluster. The example includes a LLM fine tuning script, Dockerfile, and Helm chart. |
+| `intel/intel-optimized-pytorch:2.3.0-serving-cpu` | CPU | [TorchServe* with Kubernetes] | Demonstrates using TorchServe* with Intel® Xeon® Scalable Processors to serve models on multinodes nodes from a Kubernetes cluster. The example includes a Helm chart. |
+
+## Build from Source
+
+To build the images from source, clone the [Intel® AI Containers] repository, follow the main `README.md` file to setup your environment, and run the following command:
+
+```bash
+cd workflows/charts/huggingface-llm
+docker compose build huggingface-llm
+docker compose run huggingface-llm sh -c "python /workspace/scripts/finetune.py --help"
+```
+
+## License
+
+View the [License](https://github.com/intel/ai-containers/blob/main/LICENSE) for the [Intel® AI Containers].
+
+The images below also contain other software which may be under other licenses (such as Pytorch*, Jupyter*, Bash, etc. from the base).
+
+It is the image user's responsibility to ensure that any use of The images below comply with any relevant licenses for all software contained within.
+
+\* Other names and brands may be claimed as the property of others.
+
+<!--Below are links used in these document. They are not rendered: -->
+
+[Intel® AI Containers]: https://github.com/intel/ai-containers
+[Distributed LLM Fine Tuning with Kubernetes]: https://github.com/intel/ai-containers/tree/main/workflows/charts/huggingface-llm
+[TorchServe* with Kubernetes]: https://github.com/intel/ai-containers/tree/main/workflows/charts/torchserve

workflows/charts/huggingface-llm/docker-compose.yaml

@@ -23,8 +23,13 @@ services:
         BASE_IMAGE_TAG: ${BASE_IMAGE_TAG:-2.3.0-pip-multinode}
       context: .
       labels:
+        dependency.apt.google-perftools: true
+        dependency.apt.libjemalloc: true
+        dependency.apt.libomp-dev: true
+        dependency.apt.numactl: true
+        dependency.python: ${PYTHON_VERSION:-3.10}
         dependency.python.pip: requirements.txt
-        docs: pytorch
+        docs: genai
         org.opencontainers.base.name: "intel/intel-optimized-pytorch:${IPEX_VERSION:-2.3.0}-pip-multinode"
         org.opencontainers.image.name: "intel/ai-workflows"
         org.opencontainers.image.title: "Intel® Extension for PyTorch with Hugging Face LLM fine tuning"