Answer-first summary for fast verification
Answer: Configure a Vertex AI batch prediction job to apply the model to the historical data in BigQuery.
The best approach to perform inference on a historical dataset stored in BigQuery is to set up a Vertex AI batch prediction job that directly accesses the data in BigQuery. This approach leverages the scalability of Google Cloud infrastructure and avoids unnecessary data movement or export operations. Options A and C involve exporting data to Cloud Storage in either Avro or CSV formats, which incites additional storage costs and overhead. While Option B involves importing the TensorFlow model using the CREATE MODEL statement in BigQuery ML, it might have limitations with certain model architectures or features, and it is not as optimized for large-scale batch inference compared to Vertex AI. Therefore, Option D is the most efficient and practical solution.
Author: LeetQuiz Editorial Team
Ultimate access to all questions.
You are tasked with implementing a batch inference machine learning pipeline using Google Cloud services. The machine learning model, developed using TensorFlow, is stored in the SavedModel format in Cloud Storage. Your objective is to apply this model to a historical dataset, which contains 10 TB of data stored in a BigQuery table. Considering the large size of the dataset and the need for efficient processing, how should you perform the inference?
A
Export the historical data to Cloud Storage in Avro format. Configure a Vertex AI batch prediction job to generate predictions for the exported data.
B
Import the TensorFlow model by using the CREATE MODEL statement in BigQuery ML. Apply the historical data to the TensorFlow model.
C
Export the historical data to Cloud Storage in CSV format. Configure a Vertex AI batch prediction job to generate predictions for the exported data.
D
Configure a Vertex AI batch prediction job to apply the model to the historical data in BigQuery.
No comments yet.