Professional Machine Learning Engineer

Question

You developed a Vertex AI pipeline that trains a classification model on data stored in a large BigQuery table. The pipeline has four steps, each created by a Python function that uses the KubeFlow v2 API. The components have the following names:

Question Image

You launch the Vertex AI pipeline as follows:

Question Image

You perform many model iterations by adjusting the code and parameters of the training step. You observe high development costs, particularly for the data-export and preprocessing steps. You need to reduce model-development costs. What should you do?

Explanation

Vertex AI Pipelines reuses cached outputs only when a step’s inputs, output definition, and component specification match within the same pipeline name. Stable component names for the unchanged export and preprocessing components allow their prior outputs to be reused, while a timestamped training component ensures changed training code is rerun. This avoids repeatedly exporting the large BigQuery table and preprocessing its data.

Learn more

Configure execution caching | Vertex AI

Professional Machine Learning Engineer

AstroTutor

Notes

Readiness

QuestionQ49

Community Discussion