Answer-first summary for fast verification
Answer: choice
Grid sampling in Azure Machine Learning can only be used with discrete hyperparameters, specifically those defined using the 'choice' sampling method. The 'choice' method allows you to specify a list of discrete values for hyperparameters like num_hidden_layers and batch_size, and grid sampling will then evaluate all possible combinations of these values. This is explicitly stated in Microsoft's documentation and confirmed by the community discussion where all comments (with upvotes ranging from 1 to 6) unanimously select option C, citing that grid sampling is only compatible with choice hyperparameters. Other options like uniform, qlognormal, and normal are continuous sampling methods that cannot be used with grid sampling.
Author: LeetQuiz Editorial Team
Ultimate access to all questions.
You are implementing hyperparameter tuning for a model training script from a notebook in an Azure Machine Learning workspace. You need to configure a grid sampling method over the search space for the num_hidden_layers and batch_size hyperparameters. Which hyperparameter sampling method should you use?
A
uniform
B
qlognormal
C
choice
D
normal
No comments yet.