You have trained a text classification model in TensorFlow using AI Platform. You want to use the trained model for batch predictions on text data stored in BigQuery while minimizing computational…

Question

Options

• AExport the model to BigQuery ML.
• BDeploy and version the model on AI Platform.
• CUse Dataflow with the SavedModel to read the data from BigQuery.
• DSubmit a batch prediction job on AI Platform that points to the model location in Cloud Storage.

• A
  69% (35)
• B
  4% (2)
• C
  18% (9)
• D
  10% (5)

Explanation

The goal is batch predictions on BigQuery data with minimal computational overhead. Exporting the TensorFlow model to BigQuery ML (using CREATE MODEL with a TensorFlow SavedModel import) allows predictions to run directly inside BigQuery using ML.PREDICT, meaning the data never leaves BigQuery. This eliminates the overhead of moving data to an external serving system, setting up infrastructure, or managing prediction jobs separately. Option B (deploying on AI Platform) is designed for online serving, and would require exporting data out of BigQuery first. Option C (Dataflow with SavedModel) introduces an additional processing layer and significant infrastructure overhead. Option D (AI Platform batch prediction) requires data to be exported from BigQuery to Cloud Storage before the job can run, adding overhead. BigQuery ML is the most efficient path for in-place batch inference on BigQuery data.

No community discussion yet for this question.