Skip to content

Monitoring with MLflow

Open in Colab ยท View on GitHub

Uses MLflow as the registry and shows what lands there: one run per version (params, metrics, artefacts, pyfunc model), a state run, and decision logs.

pip install pecca mlflow requests
jupyter lab examples/notebooks/monitoring_mlflow.ipynb
Runtime: ~40 s. Needs Docker: no (local MLflow server). With PECCA_SKIP_DOCKER=1 (as in CI) the Docker cells are skipped and the notebook still runs end to end. Outputs are committed, so GitHub renders the results without running anything.