Module Quiz » Week 2 » ML Pipelines On Google Cloud

Q: Which orchestrator would you select for single machine development and debugging of your TFX pipeline?

Correct! The DirectRunner enables quicker experimentation and debugging of your pipeline locally on a single machine with less dependency and resource management overhead than Airflow or Kubeflow.

Q: Select all the following that are outputs of a TFX pipeline run on Google Cloud:

Correct! A TFX pipeline Pusher component can be configured to deploy your pipeline’s SavedModel to Cloud Storage or AI Platform Prediction for inference. <br> Correct! Pipeline logs generated during pipeline runs executed on a Kubeflow cluster are written to Cloud Logging. <br> Correct! TFX manages consistent pipeline directory organization for pipeline artifacts on Cloud Storage.