Feature/dockerfile (#10)

* add stuff

* add more stuff

* update eval, deps

* update readme

.dockerignore

@@ -0,0 +1,163 @@
+# IDE stuff
+.vscode/
+
+# 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/

.env.example

@@ -0,0 +1,2 @@
+RAW_DATA_ROOT_PATH=/path/to/dir
+TRAIN_ARTIFACTS_ROOT_PATH=/path/to/dir

Dockerfile

@@ -0,0 +1,49 @@
+FROM python:3.11-slim as build
+
+# install poetry - respects $POETRY_VERSION & $POETRY_HOME
+# The --mount will mount the buildx cache directory to where
+# Poetry and Pip store their cache so that they can re-use it
+
+ENV PYTHONUNBUFFERED=1 \
+    # pip
+    PIP_DISABLE_PIP_VERSION_CHECK=on \
+    PIP_DEFAULT_TIMEOUT=100 \
+    POETRY_HOME="/opt/poetry" \
+    POETRY_NO_INTERACTION=1 \
+    POETRY_VIRTUALENVS_CREATE=false \
+    PATH="$POETRY_HOME/bin:$PATH"
+
+
+RUN apt-get update && \
+    apt-get install -y \
+    apt-transport-https \
+    gnupg \
+    ca-certificates \
+    build-essential \
+    git \
+    nano \
+    curl
+
+RUN curl -sSL https://install.python-poetry.org | python3 -
+
+
+# used to init dependencies
+COPY poetry.lock pyproject.toml README.md ./
+COPY ml_pipelines ./ml_pipelines
+RUN PATH=$POETRY_HOME/bin:$PATH poetry build
+
+# # # Stage 2: Production stage
+FROM python:3.11-slim
+
+ARG UID=1000
+ARG GID=1000
+WORKDIR /package
+
+RUN  groupadd -g "${GID}" python \
+  && useradd --create-home --no-log-init -u "${UID}" -g "${GID}" python \
+  && chown python:python -R /package
+
+COPY --from=build /dist/*.whl /package/dist/.
+RUN pip install /package/dist/*.whl
+
+USER python

README.md

@@ -9,9 +9,22 @@ Simple ML pipeline repo for experimenting with CI/CD / DevOps / MLOps.
 ~~- add loggers to stuff~~
 ~~- add local deployment code...~~
 ~~- add versioning to training... in deployment?~~
-- add eval pipeline, model comparison
-- add "best model" mark. add "get_best_model"
+~~- add eval pipeline, model comparison~~
+~~- add "best model" mark. add "get_best_model"~~
+~~- add Dockerfile~~
+- add real prediction logging func
 - add db conn / func to save inference cases (local deployment)
-- add Dockerfile
 - add build script to push to ECR (AWS deployment)
 - add rest of AWS deployment (using S3, EC2, AWS CodePipeline)
+
+# Commands to remember
+- python ml_pipelines/deployment/local/train.py
+- python ml_pipelines/deployment/local/eval.py
+- python ml_pipelines/deployment/local/serve.py
+- python -m ml_pipelines train_local
+- python -m ml_pipelines eval_local
+- python -m ml_pipelines serve_local
+- sudo docker build -t ml_pipelines:latest .
+- sudo docker run --rm -it ml_pipelines:latest /bin/sh
+- sudo docker run --rm -it -v localpath:containerpath --env-file .env ml_pipelines:latest /bin/bash -c "python -m ml_pipelines serve_local"
+- sudo docker run --rm -it -v localpath:containerpath --env-file .env --network host ml_pipelines:latest /bin/bash -c "python -m ml_pipelines serve_local"

ml_pipelines/__main__.py

@@ -0,0 +1,13 @@
+import typer
+
+from ml_pipelines.deployment.local.eval import main as eval_local
+from ml_pipelines.deployment.local.serve import main as serve_local
+from ml_pipelines.deployment.local.train import main as train_local
+
+app = typer.Typer()
+
+app.command("train_local")(train_local)
+app.command("eval_local")(eval_local)
+app.command("serve_local")(serve_local)
+
+app()

ml_pipelines/deployment/local/common.py

@@ -100,13 +100,17 @@ def get_latest_version(root_path: os.PathLike, filename: str) -> str:
     root_dir = Path(root_path)
     versions: list[tuple[str, float]] = []
     for version_dir in root_dir.iterdir():
+        if not version_dir.is_dir():
+            continue
         st_mtime = (version_dir / filename).stat().st_mtime
         versions.append((version_dir.stem, st_mtime))
     return max(versions, key=lambda t: t[1])[0]
 
 
 def get_best_version(train_artifacts_root_path: os.PathLike):
     train_dir = Path(train_artifacts_root_path)
+    if "best_model" not in set(f for f in train_dir.iterdir() if f.is_file()):
+        return None
     with open(train_dir / "best_model") as f:
         return f.read()
 

ml_pipelines/deployment/local/eval.py

@@ -21,6 +21,7 @@ def run_eval_comparison_pipeline(  # noqa: PLR0913
     raw_data_root_path: os.PathLike,
     train_versions: list[str],
     train_artifacts_root_path: os.PathLike,
+    tag_best_model: bool,
     logger: Logger,
 ):
     logger.info(f"Running eval pipeline on model versions: {train_versions}.")
@@ -39,43 +40,66 @@ def run_eval_comparison_pipeline(  # noqa: PLR0913
         )
         all_metrics.append((v, metrics))
     best_version = max(all_metrics, key=lambda t: t[1])[0]
-    logger.info(f"Tagging best version as {best_version}")
-    tag_best_version(best_version, train_artifacts_root_path)
+    if tag_best_model and len(train_versions) > 1:
+        logger.info(f"Tagging best version as {best_version}")
+        tag_best_version(best_version, train_artifacts_root_path)
 
 
-if __name__ == "__main__":
-    from dotenv import load_dotenv
-
-    load_dotenv()
-    RAW_DATA_ROOT_DIR = os.environ["RAW_DATA_ROOT_DIR"]
-    TRAIN_ARTIFACTS_ROOT_DIR = os.environ["TRAIN_ARTIFACTS_ROOT_DIR"]
-
-    def main(
-        raw_data_version: Union[str, None] = None,  # noqa: UP007
-        train_versions: Union[list[str], None] = None,  # noqa: UP007
-        raw_data_root_path: str = RAW_DATA_ROOT_DIR,
-        train_artifacts_root_path: str = TRAIN_ARTIFACTS_ROOT_DIR,
-    ):
-        logger = Logger(__file__)
-        logger.addHandler(logging.StreamHandler(sys.stdout))
-
-        if raw_data_version is None:
-            raw_data_version = get_latest_version(
-                raw_data_root_path,  # type: ignore
-                "raw_data.csv",
-            )
-
-        if not train_versions:
-            train_versions = get_all_available_train_versions(  # type: ignore
-                train_artifacts_root_path
-            )
-
-        run_eval_comparison_pipeline(  # noqa: PLR0913
-            raw_data_version=raw_data_version,
-            raw_data_root_path=raw_data_root_path,  # type: ignore
-            train_versions=train_versions,  # type: ignore
-            train_artifacts_root_path=train_artifacts_root_path,  # type: ignore
-            logger=logger,
+def main(
+    raw_data_root_path: Union[str, None] = None,  # noqa: UP007
+    train_artifacts_root_path: Union[str, None] = None,  # noqa: UP007, E501
+    raw_data_version: Union[str, None] = None,  # noqa: UP007
+    train_versions: Union[list[str], None] = None,  # noqa: UP007
+    tag_best_model: bool = False,
+):
+    """
+    Runs the model evaluation and comparison pipeline using local paths
+    for inputs and outputs.
+
+    If `raw_data_root_path` is null, the command searches for the RAW_DATA_ROOT_PATH
+    environment variable, and if not present, assumes this to be "/".
+
+    If `train_artifacts_root_path` is null, the command searches for the
+    TRAIN_ARTIFACTS_ROOT_PATH environment variable, and if not present,
+    assumes this to be "/".
+
+    If `raw_data_version` is null, the command searches for the latest version in
+    `raw_data_root_path`.
+
+    If `train_versions` is null or empty, the command automatically evaluates all
+    models found in `train_artifacts_root_path`.
+
+    If `tag_best_model` is set (to true) and more than one model version is evaluated,
+    the best performing one is tagged as the best version.
+    """
+    logger = Logger(__file__)
+    logger.addHandler(logging.StreamHandler(sys.stdout))
+
+    if raw_data_root_path is None:
+        raw_data_root_path = os.environ.get("RAW_DATA_ROOT_PATH", "/")
+    if train_artifacts_root_path is None:
+        train_artifacts_root_path = os.environ.get("TRAIN_ARTIFACTS_ROOT_PATH", "/")
+
+    if raw_data_version is None:
+        raw_data_version = get_latest_version(
+            raw_data_root_path,  # type: ignore
+            "raw_data.csv",
+        )
+
+    if not train_versions:
+        train_versions = get_all_available_train_versions(  # type: ignore
+            train_artifacts_root_path  # type: ignore
         )
 
+    run_eval_comparison_pipeline(  # noqa: PLR0913
+        raw_data_version=raw_data_version,
+        raw_data_root_path=raw_data_root_path,  # type: ignore
+        train_versions=train_versions,  # type: ignore
+        train_artifacts_root_path=train_artifacts_root_path,  # type: ignore
+        tag_best_model=tag_best_model,
+        logger=logger,
+    )
+
+
+if __name__ == "__main__":
     typer.run(main)