You manage an Azure Machine Learning workspace. An MLflow model is already registered. You plan to customize how the deployment does inference. You need to deploy the MLflow model to a batch…

The correct answer is A. scoring script. When customizing how an MLflow model performs inference on a batch endpoint, the first thing you must create is a scoring script. By default, Azure ML uses the MLflow model's built-in inference logic, but to override or customize that behavior you must supply a custom scoring…

Train and deploy models

Question

You manage an Azure Machine Learning workspace. An MLflow model is already registered. You plan to customize how the deployment does inference. You need to deploy the MLflow model to a batch endpoint for batch inferencing. What should you create first?

Options

Ascoring script
Bdeployment
Cenvironment
Ddeployment definition

How the community answered

(18 responses)

A
89% (16)
C
6% (1)
D
6% (1)

Explanation

When customizing how an MLflow model performs inference on a batch endpoint, the first thing you must create is a scoring script. By default, Azure ML uses the MLflow model's built-in inference logic, but to override or customize that behavior you must supply a custom scoring script that defines the init() and run() functions. The scoring script must exist before you can create the deployment (or deployment definition) that references it, and before the environment (which packages the dependencies) is built around it.

Topics

#Azure ML Deployment#Batch Inference#Scoring Script#MLflow Models

Community Discussion

No community discussion yet for this question.

Full DP-100 Practice