Databricks Certified Machine Learning Professional — Question 63

A machine learning engineer is converting a Hyperopt-based hyperparameter tuning process from manual MLflow logging to MLflow Autologging. They notice that not all details and objects are automatically logged, and they will need to manually log some things.

Which of the following will need to be manually logged when performing nested runs with Hyperopt and MLflow Autologging?

Answer options

• A. Trial models
• B. Trial status
• C. Best trial evaluation metric
• D. Hyperparameter values
• E. Evaluation metrics

Correct answer: C

Explanation

The best trial evaluation metric needs to be manually logged because MLflow Autologging does not capture this specific detail automatically during nested runs with Hyperopt. Other options such as trial models, trial status, hyperparameter values, and evaluation metrics are typically handled by the autologging feature.