You create a script that trains a convolutional neural network model over multiple epochs and logs the validation loss after each epoch. The script includes arguments for batch size and learning…

Question

Options

• ARun the script in an experiment based on an AutoMLConfig object
• BCreate a PythonScriptStep object for the script and run it in a pipeline
• CUse the Automated Machine Learning interface in Azure Machine Learning studio
• DRun the script in an experiment based on a ScriptRunConfig object
• ERun the script in an experiment based on a HyperDriveConfig object

• A
  14% (5)
• B
  3% (1)
• C
  3% (1)
• D
  8% (3)
• E
  73% (27)

Explanation

HyperDriveConfig is the correct tool for hyperparameter tuning in Azure ML. It wraps a ScriptRunConfig with a parameter search space (e.g., discrete choices for batch size and learning rate), a primary metric to optimize (in this case, minimize validation loss), a sampling strategy (random, grid, or Bayesian), and early termination policies. Azure ML then orchestrates multiple child runs across the compute cluster, each testing a different hyperparameter combination, and identifies the run with the lowest validation loss. AutoMLConfig (A) and the AutoML Studio interface (C) automate algorithm and feature selection - not hyperparameter tuning of a custom script. PythonScriptStep (B) is a pipeline step with no tuning capability. ScriptRunConfig (D) runs the script once with fixed parameters, with no search.

No community discussion yet for this question.