Databricks Machine Learning Associate Question 465
Select 4A data science team is using MLflow to track their machine learning experiments in Databricks. They want to compare the performance of different models and analyze their parameters. Which of the following pieces of information can they view in the MLflow UI?
- A
Logged metrics such as accuracy or loss
- B
Hyperparameters used in the experiments
- C
The code version or Git commit hash used for the experiment
- D
The exact GPU specifications used during training
- E
The artifacts generated during the experiment, such as model files
Show answer and explanation
Correct answers: A, B, C, E
Explanation
The MLflow UI is a comprehensive interface for tracking and analyzing machine learning experiments. It allows users to track metrics, parameters, code versions, and artifacts, providing a centralized view of all the key components of an experiment. However, it does not automatically track hardware specifications like GPUs unless explicitly logged.
- A. Correct.
The MLflow UI allows users to view logged metrics like accuracy, loss, and other evaluation metrics tracked during the experiment.
- B. Correct.
Hyperparameters (parameters used to train the model) are logged as parameters in MLflow and can be viewed in the MLflow UI.
- C. Correct.
The MLflow UI can display the code version or Git commit hash if this information is logged as part of the experiment.
- D. Incorrect.
While hardware specifications can be logged manually, MLflow does not automatically track or display specific GPU details in the UI.
- E. Correct.
Artifacts such as model files, plots, and other outputs from the experiment are tracked and accessible in the MLflow UI.