Amazon SageMaker Automatic Model Tuning now supports Grid Search
Amazon SageMaker Automatic Model Tuning now supports Grid Search to enable use cases that require reproducibility of hyperparameter tuning. Grid search will cover every combination of the specified hyperparameter values and yield reproducible tuning results.
Amazon SageMaker Automatic Model Tuning now supports Grid Search to enable use cases that require reproducibility of hyperparameter tuning. Grid search will cover every combination of the specified hyperparameter values and yield reproducible tuning results.
Amazon SageMaker Automatic Model Tuning allows you to tune and find the most accurate version of a machine learning model by searching for the optimal set of hyperparameter configurations for your dataset using various search strategies. Before this launch, you had the option to tune your models through "Random", "Bayesian" or "Hyperband" search strategies. Starting today, you can choose Grid search for hyperparameter optimization. When compared to "Random", "Bayesian" or "Hyperband", Grid search determines which regions of the hyperparameter search space are most promising by exhaustively exploring each and every combination of the specified hyperparameters. This makes grid search the preferred choice for use cases where reproducibility of hyperparameter tuning is important.
Grid Search is now available for SageMaker Automatic Model Tuning in all commercial AWS Regions. To learn more, review the blog post or visit SageMaker Automatic Model Tuning web page.