Skip to content

00_environment_prerequisites.md

Environment Prerequisites

Before starting the Bike Demand Forecasting Pipeline lab, ensure your environment is correctly configured.

1. Required Software

  • Python 3.11+
  • Kubeflow Pipelines (KFP) v2
  • MLflow Tracking Server
  • S3-compatible storage (for pipeline artifacts)

2. Environment Variables

The pipeline relies on several environment variables for configuration. Ensure these are set in your pipeline runtime:

  • DATASET_URL: The URL of the UCI bike sharing dataset.
  • MLFLOW_REMOTE_TRACKING_SERVER: The URI of your MLflow tracking server (e.g., http://mlflow-tracking.mlflow.svc.cluster.local:80).
  • PARTICIPANT_FIRSTNAME: Your first name, used to namespace experiments and avoid conflicts with other participants.

3. Dependencies

If you are performing local testing of components, install the required dependencies:

pip install -r labs/02_intermediate/bike_demand_forecasting_pipeline/requirements.txt