DP-100 Question #303: Real Exam Question with Answer & Explanation

Question

You use the Azure Machine Learning designer to create and run a training pipeline. The pipeline must be run every night to inference predictions from a large volume of files. The folder where the files will be stored is defined as a dataset. You need to publish the pipeline as a REST service that can be used for the nightly inferencing run. What should you do?

Options

• ACreate a batch inference pipeline
• BSet the compute target for the pipeline to an inference cluster
• CCreate a real-time inference pipeline
• DClone the pipeline

Explanation

{"question_number": 10, "correct_answer": "A", "explanation": "The correct answer is A: Create a batch inference pipeline. The scenario describes nightly processing of a large volume of files stored in a folder dataset-a classic batch inferencing workload. In Azure ML Designer, you create a batch inference pipeline (distinct from a real-time inference pipeline) by cloning the training pipeline and adding batch scoring components. This can then be published as a REST endpoint that can be triggered on a schedule. Option B is wrong because inference clusters are for real-time (AKS) deployments. Option C (real-time inference pipeline) is for online, low-latency single-request predictions, not large file volumes. Option D (clone the pipeline) alone is incomplete-cloning is a step within the process, not the solution itself.", "generated_by": "claude-sonnet", "llm_judge_score": 4}

No community discussion yet for this question.