Amazon SageMaker - AWS GovCloud (US)

Amazon SageMaker

This service is currently available in AWS GovCloud (US-West) only.

Amazon SageMaker is a fully managed machine learning service. With Amazon SageMaker, data scientists and developers can quickly and easily build and train machine learning models, and then directly deploy them into a production-ready hosted environment. It provides an integrated Jupyter authoring notebook instance for easy access to your data sources for exploration and analysis, so you don't have to manage servers. It also provides common machine learning algorithms that are optimized to run efficiently against extremely large data in a distributed environment. With native support for bring-your-own-algorithms and frameworks, Amazon SageMaker provides flexible distributed training options that adjust to your specific workflows.

How Amazon SageMaker Differs for AWS GovCloud (US)

  • The following instance types are not supported in the AWS GovCloud (US) Region: t3.[medium, large, xlarge, 2xlarge] and p2.[xlarge, 8xlarge, 16xlarge].

  • The following services are not available in AWS GovCloud (US): Code Repositories, SageMaker Studio, SageMaker Environments, SageMaker GroundTruth, SageMaker Marketplace, SageMaker Search, SageMaker Experiments, and SageMaker Reinforcement Learning.

  • The associated API calls for these services are available but will fail with a 4xx message indicating “The requested operation is not supported in the called region".

  • The following features are not available in AWS GovCloud (US):

    • AWS Marketplace for SageMaker Models and Algorithms

    • Elastic Inference

    • SageMaker Algorithm resources

    • SageMaker Clarify

    • SageMaker Code Repositories

    • SageMaker Data Wrangler

    • SageMaker Distributed

    • SageMaker Edge Manager

    • SageMaker Environments

    • SageMaker Feature Store

    • SageMaker GroundTruth

    • SageMaker JumpStart

    • SageMaker Neo

    • SageMaker Pipelines

    • SageMaker Reinforcement Learning

    • SageMaker Search

    • SageMaker Studio

Documentation for Amazon SageMaker

Amazon SageMaker documentation.

ITAR Boundary

AWS GovCloud (US) has an ITAR boundary, which defines where customers are allowed to store ITAR-controlled data for this service in AWS GovCloud (US) Regions. To maintain ITAR compliance, you must place ITAR-controlled data on the applicable part of the ITAR boundary. If you do not have any ITAR-controlled data in AWS GovCloud (US) Regions, this section does not apply to you. The following information identifies the ITAR boundary for this service:

ITAR-Regulated Data Permitted ITAR-Regulated Data Not Permitted
  • All data entered, stored, and processed within a Notebook Instance and ephemeral drives can contain ITAR-regulated data. All data processed during training, automatic model tuning, batch transformation, and endpoint invocation can contain ITAR-regulated data.

  • Amazon SageMaker metadata is not permitted to contain ITAR-regulated data. This metadata includes all configuration data that you enter when creating and maintaining your NotebookInstances, NotebookInstanceLifecycleConfigs, Endpoints, Models, EndpointConfigs, TrainingJobs, HyperParameterTuningJobs, and BatchTransformJobs.

    Do not enter ITAR-regulated data in the following console fields:

    • NotebookInstance Name

    • NotebookInstanceLifecycleConfig Name

    • Model Name

    • Model Container Hostname

    • Model Environment names and values

    • Endpoint Name

    • Endpoint Config Name

    • Endpoint Config Production Variant names

    • Endpoint Config

    • TrainingJob Name

    • BatchTransformJob Name

    • Hyperparameter Names or values

    • Input Channel Name

    • Any resource tag or value.

    • Names of any metrics emitted by algorithms.

    • Names of any training or inference container environment variables.