You have developed a custom ML model using Vertex AI and want to deploy it for online serving. You need to optimize the model's serving performance by ensuring that the model can handle high…

Question

Options

• ADeploy the model to a Vertex AI endpoint resource to automatically scale the serving backend
• BImplement a containerized serving solution using Cloud Run. Configure the concurrency settings
• CApply simplification techniques such as model pruning and quantization to reduce the model's size
• DEnable request-response logging for the model hosted in Vertex AI. Use Looker Studio to analyze

• A
  80% (40)
• B
  4% (2)
• C
  2% (1)
• D
  14% (7)

Explanation

The requirement is high-throughput, low-latency online serving with the simplest solution. Vertex AI endpoints provide a fully managed online prediction service with built-in autoscaling-as traffic increases, Vertex AI automatically provisions additional serving replicas to maintain low latency and handle high throughput, with no configuration required beyond specifying machine type. Option B (Cloud Run) requires manually configuring concurrency, scaling parameters, and containerizing the serving logic-more complex. Option C (pruning and quantization) is a model optimization technique, not a deployment strategy, and does not directly address serving infrastructure. Option D (request-response logging with Looker Studio) is a monitoring approach, not a serving optimization. Vertex AI endpoints are the simplest managed path to scalable, low-latency online serving.

No community discussion yet for this question.