Cookiecutter template to kick off your machine learning experiments with MLflow, Jupyter and more.
The template makes it easy to kick off simple machine learning experiments (xgboost, sklearn) conducted in Jupyter notebooks and tracked with MLflow.
It contains an example notebook to train and evaluate machine learning models compatible with the sklearn API.
Tools used in the generated project:
- Poetry for dependency and environment management.
- Basic data science dependencies are included (see in the pyproject.toml)
- Omegaconf for configuration management
- Jupyter for running notebooks
- Mlflow to track experiments
.envfiles for configuring environment variables
- Python 3
- cookiecutter (tested with 2.3, see installation notes here)
- cruft (tested with 2.15, see installation notes here)
cruft create [email protected]:julcsii/simple-ml-template.git
cruft update
cruft update --variables-to-update '{ "<variable>": "<new value>" }'