Document History for Amazon SageMaker - Amazon SageMaker

Document History for Amazon SageMaker

Change Description Date

Studio Notebooks

SageMaker Studio Notebooks

April 28, 2020

New features re:Invent 2019

SageMaker Studio, SageMaker Studio Notebooks (preview), SageMaker Experiments, SageMaker Autopilot, SageMaker Debugger, SageMaker Model Monitor

December 3, 2019

New features re:Invent 2018

Amazon SageMaker Ground Truth, Using Elastic Inference in Amazon SageMaker, SageMaker Resources in AWS Marketplace, SageMaker Inference Pipelines, SageMaker Neo, Manage Machine Learning Experiments with Search , Use Reinforcement Learning in Amazon SageMaker, Associating Git Repositories with Amazon SageMaker Notebook Instances, Semantic Segmentation, Using Augmented Manifest Files in TrainingJobs

November 28, 2018

Configuring notebook instances

You can use shell scripts to configure notebook instances when you create or start them. For more information, see Customize a Notebook Instance.

May 1, 2018

Disable direct internet access

You can now disable direct internet access for notebook instances. For more information, see Notebook Instances Are Enabled with Internet Access by Default.

March 15, 2018

Application Auto Scaling support

Amazon SageMaker now supports Application Auto Scaling for production variants. For information, see Automatically Scaling SageMaker Models

February 28, 2018

TensorFlow 1.5 and MXNet 1.0 support

Amazon SageMaker Deep Learning containers now support TensorFlow 1.5 and Apache MXNet 1.0.

February 27, 2018

BlazingText algorithm

Amazon SageMaker now supports the BlazingText algorithm.

January 18, 2018

KMS encryption support for training and hosting

Amazon SageMaker now supports KMS encryption for hosting instances and training model artifacts at rest. You can specify a AWS Key Management Service key that Amazon SageMaker uses to encrypt data on the storage volume attached to a hosting endpoint by using the KmsKeyId request parameter in a call to CreateEndpointConfig. You can specify an AWS KMS key that SageMaker uses to encrypt training model artifacts at rest by setting the KmsKeyId field of the OutputDataConfig object you use to configure your training job.

January 17, 2018

CloudTrail support

Amazon SageMaker now supports logging with AWS CloudTrail.

January 11, 2018

DeepAR Forecasting algorithm

Amazon SageMaker now supports the DeepAR algorithm for time series forecasting.

January 8, 2018