Among the general flurry of announcements to have emerged from Amazon Web Services’ (AWS) 2020 re:Invent conference is the news that the cloud vendor’s SageMaker machine learning (ML) service has been imbued with a bunch of new capabilities.
According to Amazon, the new capabilities will work to make it easier for developers to automate and scale all steps of the end-to-end machine learning workflow.
Broadly, AWS claims to have brought together new capabilities such as faster data preparation, a purpose-built repository for prepared data, workflow automation, greater transparency into training data to mitigate bias and explain predictions, distributed training capabilities to train large models up to two times faster and model monitoring on edge devices.
The new capabilities are:
Amazon SageMaker Data Wrangler — providing a faster and easier way for developers to prepare data for machine learning.
Amazon SageMaker Feature Store — delivering a purpose-built data store for storing, updating, retrieving, and sharing machine learning features.
Amazon SageMaker Pipelines — giving developers what AWS claims is the first purpose-built, easy-to-use continuous integration and continuous delivery (CI/CD) service for machine learning.
Amazon SageMaker Clarify — providing developers with greater visibility into their training data so they can limit bias in machine learning models and explain predictions.
Deep profiling for Amazon SageMaker Debugger — to monitor machine learning training performance to help developers train models faster.
Distributed Training on Amazon SageMaker — delivering new capabilities that can train large models up to two times faster than would otherwise be possible with today’s machine learning processors.
Amazon SageMaker Edge Manager — delivering machine learning model monitoring and management for edge devices to ensure that models deployed in production are operating correctly.
Amazon SageMaker JumpStart — providing a developer portal for pre-trained models and pre-built workflows.