Amazon SageMaker Automatic Model Tuning: Scalable Gradient-Free Optimization

TL;DR

Amazon SageMaker Automatic Model Tuning (AMT) is a scalable, fully managed system that automates hyperparameter optimization for machine learning models using gradient-free methods like random search and Bayesian optimization.

Contribution

The paper introduces AMT, a comprehensive system for scalable, gradient-free hyperparameter tuning that integrates with various machine learning frameworks and includes advanced features like early stopping.

Findings

AMT effectively finds optimal hyperparameters improving model performance.

Automated early stopping reduces tuning time and computational costs.

Warm-starting enhances tuning efficiency in iterative experiments.

Abstract

Tuning complex machine learning systems is challenging. Machine learning typically requires to set hyperparameters, be it regularization, architecture, or optimization parameters, whose tuning is critical to achieve good predictive performance. To democratize access to machine learning systems, it is essential to automate the tuning. This paper presents Amazon SageMaker Automatic Model Tuning (AMT), a fully managed system for gradient-free optimization at scale. AMT finds the best version of a trained machine learning model by repeatedly evaluating it with different hyperparameter configurations. It leverages either random search or Bayesian optimization to choose the hyperparameter values resulting in the best model, as measured by the metric chosen by the user. AMT can be used with built-in algorithms, custom algorithms, and Amazon SageMaker pre-built containers for machine learning frameworks. We discuss the core functionality, system architecture, our design principles, and lessons learned. We also describe more advanced features of AMT, such as automated early stopping and warm-starting, showing in experiments their benefits to users.