Databricks Machine Learning Associate Question 647
Select 3A data science team has trained a machine learning model in Databricks and wants to deploy it for real-time inference. They decide to use Databricks Model Serving. Which of the following steps are required to deploy and query the model for real-time inference?
- A
Register the trained model in the Databricks Model Registry.
- B
Enable Model Serving for the registered model.
- C
Use the Databricks REST API or client libraries to query the deployed model.
- D
Export the model to an external service before deploying it for real-time inference.
- E
Manually set up a web server to host the model for real-time serving.
Show answer and explanation
Correct answers: A, B, C
Explanation
To deploy and query a model for real-time inference in Databricks, the model must first be registered in the Databricks Model Registry. Enabling Model Serving is then required to deploy the model. Once deployed, real-time inference can be achieved using the Databricks REST API or client libraries. Databricks Model Serving manages hosting and deployment internally, so there is no need to export the model or set up a web server manually.
- A. Correct.
Correct. Registering the model in the Databricks Model Registry is a required step to enable Model Serving in Databricks.
- B. Correct.
Correct. Enabling Model Serving for the registered model is necessary to deploy it for real-time inference in Databricks.
- C. Correct.
Correct. Querying the deployed model for real-time inference can be done using the Databricks REST API or client libraries.
- D. Incorrect.
Incorrect. Databricks Model Serving does not require exporting the model to an external service; it manages deployment within the Databricks environment.
- E. Incorrect.
Incorrect. Databricks Model Serving automatically handles hosting, so there is no need to manually set up a web server.