دسترسی نامحدود
برای کاربرانی که ثبت نام کرده اند
برای ارتباط با ما می توانید از طریق شماره موبایل زیر از طریق تماس و پیامک با ما در ارتباط باشید
در صورت عدم پاسخ گویی از طریق پیامک با پشتیبان در ارتباط باشید
برای کاربرانی که ثبت نام کرده اند
درصورت عدم همخوانی توضیحات با کتاب
از ساعت 7 صبح تا 10 شب
ویرایش:
نویسندگان: coll
سری:
ناشر: Amazon Web Services
سال نشر: 2021
تعداد صفحات: [2621]
زبان: English
فرمت فایل : PDF (درصورت درخواست کاربر به PDF، EPUB یا AZW3 تبدیل می شود)
حجم فایل: 66 Mb
در صورت تبدیل فایل کتاب Amazon SageMaker Developer Guide به فرمت های PDF، EPUB، AZW3، MOBI و یا DJVU می توانید به پشتیبان اطلاع دهید تا فایل مورد نظر را تبدیل نمایند.
توجه داشته باشید کتاب راهنمای توسعه دهنده Amazon SageMaker نسخه زبان اصلی می باشد و کتاب ترجمه شده به فارسی نمی باشد. وبسایت اینترنشنال لایبرری ارائه دهنده کتاب های زبان اصلی می باشد و هیچ گونه کتاب ترجمه شده یا نوشته شده به فارسی را ارائه نمی دهد.
Amazon SageMaker Table of Contents What Is Amazon SageMaker? Amazon SageMaker Features Amazon SageMaker Pricing Are You a First-time User of Amazon SageMaker? How Amazon SageMaker Works Machine Learning with Amazon SageMaker Explore, Analyze, and Process Data What Is Fairness and Model Explainability for Machine Learning Predictions? Best Practices for Evaluating Fairness and Explainability in the ML Lifecycle Sample Notebooks Guide to the SageMaker Clarify Documentation Train a Model with Amazon SageMaker Deploy a Model in Amazon SageMaker Deploy a Model on SageMaker Hosting Services Best Practices for Deploying Models on SageMaker Hosting Services Get Inferences for an Entire Dataset with Batch Transform Validate a Machine Learning Model Monitoring a Model in Production Use Machine Learning Frameworks, Python, and R with Amazon SageMaker Use Apache MXNet with Amazon SageMaker What do you want to do? Use Apache Spark with Amazon SageMaker Download the SageMaker Spark Library Integrate Your Apache Spark Application with SageMaker Example 1: Use Amazon SageMaker for Training and Inference with Apache Spark Use Custom Algorithms for Model Training and Hosting on Amazon SageMaker with Apache Spark Use the SageMakerEstimator in a Spark Pipeline SDK examples: Use Amazon SageMaker with Apache Spark Use Chainer with Amazon SageMaker What do you want to do? Use Hugging Face with Amazon SageMaker Training How to run training with the Hugging Face Estimator Inference How to deploy an inference job using the Hugging Face Deep Learning Containers What do you want to do? Use PyTorch with Amazon SageMaker What do you want to do? R User Guide to Amazon SageMaker R Kernel in SageMaker Get Started with R in SageMaker Example Notebooks Use Scikit-learn with Amazon SageMaker What do you want to do? Use SparkML Serving with Amazon SageMaker Use TensorFlow with Amazon SageMaker Use TensorFlow Version 1.11 and Later What do you want to do? Use TensorFlow Legacy Mode for Versions 1.11 and Earlier Supported Regions and Quotas Request a service quota increase for SageMaker resources Get Started with Amazon SageMaker Set Up Amazon SageMaker Create an AWS Account Create an IAM Administrator User and Group Onboard to Amazon SageMaker Studio Onboard to Amazon SageMaker Studio Using Quick Start Onboard to Amazon SageMaker Studio Using AWS SSO Set Up AWS SSO for Use with Amazon SageMaker Studio Onboard to Amazon SageMaker Studio Using IAM Choose a VPC Delete an Amazon SageMaker Studio Domain Delete a SageMaker Studio Domain (Studio) Delete a SageMaker Studio Domain (CLI) SageMaker JumpStart Using JumpStart Solutions Models Text Models Vision Models Deploy a model Model Deployment Configuration Fine-Tune a Model Fine-Tuning Data Source Fine-Tuning deployment configuration Hyperparameters Training Output Next Steps Amazon SageMaker Studio Tour Get Started with Amazon SageMaker Notebook Instances Machine Learning with the SageMaker Python SDK Tutorial Overview Step 1: Create an Amazon SageMaker Notebook Instance (Optional) Change SageMaker Notebook Instance Settings (Optional) Advanced Settings for SageMaker Notebook Instances Step 2: Create a Jupyter Notebook Step 3: Download, Explore, and Transform a Dataset Load Adult Census Dataset Using SHAP Overview the Dataset Split the Dataset into Train, Validation, and Test Datasets Convert the Train and Validation Datasets to CSV Files Upload the Datasets to Amazon S3 Step 4: Train a Model Choose the Training Algorithm Create and Run a Training Job Step 5: Deploy the Model to Amazon EC2 Deploy the Model to SageMaker Hosting Services (Optional) Use SageMaker Predictor to Reuse the Hosted Endpoint (Optional) Make Prediction with Batch Transform Step 6: Evaluate the Model Evaluate the Model Deployed to SageMaker Hosting Services Step 7: Clean Up Amazon SageMaker Studio Amazon SageMaker Studio UI Overview Left sidebar File and resource browser Main work area Settings Use the Amazon SageMaker Studio Launcher Notebooks and compute resources Utilities and files Studio Entity Status Use Amazon SageMaker Studio Notebooks How Are Amazon SageMaker Studio Notebooks Different from Notebook Instances? Get Started Log In from the Amazon SageMaker console Next Steps Create or Open an Amazon SageMaker Studio Notebook Open a Studio notebook Create a Notebook from the File Menu Create a Notebook from the Launcher Use the SageMaker Studio Notebook Toolbar Share and Use an Amazon SageMaker Studio Notebook Share a Notebook Use a Shared Notebook Get Notebook and App Metadata Get Notebook Metadata Get App Metadata Get Notebook Differences Get the Difference Between the Last Checkpoint Get the Difference Between the Last Commit Manage Resources Change an Instance Type Change a Kernel Shut Down Resources Shut Down an Open Notebook Shut Down Resources Usage Metering Available Resources Available SageMaker Studio Instance Types Available Amazon SageMaker Images Available Amazon SageMaker Kernels Bring your own SageMaker image Create a custom SageMaker image (Console) Attach a custom SageMaker image (Control Panel) Attach an existing image version to your domain Detach a custom SageMaker image Launch a custom SageMaker image in SageMaker Studio Bring your own custom SageMaker image tutorial Add a Studio-compatible container image to Amazon ECR Create a SageMaker image from the ECR container image Attach the SageMaker image to a new domain Attach the SageMaker image to your current domain View the attached image in the Studio control panel Clean up resources Custom SageMaker image specifications Set Up a Connection to an Amazon EMR Cluster Perform Common Tasks in Amazon SageMaker Studio Upload Files to SageMaker Studio Clone a Git Repository in SageMaker Studio Stop a Training Job in SageMaker Studio Use TensorBoard in Amazon SageMaker Studio Prerequisites Set Up TensorBoardCallback Install TensorBoard Launch TensorBoard Manage Your EFS Storage Volume in SageMaker Studio Provide Feedback on SageMaker Studio Update SageMaker Studio and Studio Apps Update SageMaker Studio Update Studio Apps Amazon SageMaker Studio Pricing Troubleshooting Amazon SageMaker Studio Use Amazon SageMaker Notebook Instances Amazon Linux 2 vs Amazon Linux notebook instances AL1 Maintenance Phase Plan Available Kernels Migrating to Amazon Linux 2 Create a Notebook Instance Access Notebook Instances Update a Notebook Instance Customize a Notebook Instance Using a Lifecycle Configuration Script Lifecycle Configuration Best Practices Install External Libraries and Kernels in Notebook Instances Package installation tools Conda Pip Unsupported Notebook Instance Software Updates Control an Amazon EMR Spark Instance Using a Notebook Example Notebooks Use or View Example Notebooks in Jupyter Classic Use or View Example Notebooks in Jupyterlab Set the Notebook Kernel Associate Git Repositories with SageMaker Notebook Instances Add a Git Repository to Your Amazon SageMaker Account Add a Git Repository to Your SageMaker Account (Console) Add a Git Repository to Your Amazon SageMaker Account (CLI) Create a Notebook Instance with an Associated Git Repository Create a Notebook Instance with an Associated Git Repository (Console) Create a Notebook Instance with an Associated Git Repository (CLI) Associate a CodeCommit Repository in a Different AWS Account with a Notebook Instance Use Git Repositories in a Notebook Instance Notebook Instance Metadata Monitor Jupyter Logs in Amazon CloudWatch Logs Automate model development with Amazon SageMaker Autopilot Get started with Amazon SageMaker Autopilot Samples: Explore modeling with Amazon SageMaker Autopilot Videos: Use Autopilot to automate and explore the machine learning process Start an AutoML job with Amazon SageMaker Autopilot Review data exploration and feature engineering automated in Autopilot. Tune models to optimize performance Choose and deploy the best model Amazon SageMaker Autopilot walkthrough Tutorials: Get started with Amazon SageMaker Autopilot Create an Amazon SageMaker Autopilot experiment Amazon SageMaker Autopilot problem types Regression Binary classification Multiclass classification Model support and validation Autopilot algorithm support Autopilot cross-validation Amazon SageMaker Autopilot model deployment Amazon SageMaker Autopilot explainability Models generated by Amazon SageMaker Autopilot Amazon SageMaker Autopilot notebooks generated to manage AutoML tasks Data exploration notebook Candidate definition notebook Configure inference output in Autopilot-generated containers Inference container definitions for regression and classification problem types Select inference response for classification models Amazon SageMaker Autopilot quotas Quotas that you can increase Resource quotas API reference guide for Amazon SageMaker Autopilot SageMaker API reference Amazon SageMaker Python SDK AWS Command Line Interface (CLI) AWS SDK for Python (Boto) AWS SDK for .NET AWS SDK for C++ AWS SDK for Go AWS SDK for Java AWS SDK for JavaScript AWS SDK for PHP V3 AWS SDK for Ruby V3 Label Data Use Amazon SageMaker Ground Truth to Label Data Are You a First-time User of Ground Truth? Getting started Step 1: Before You Begin Next Step 2: Create a Labeling Job Next Step 3: Select Workers Next Step 4: Configure the Bounding Box Tool Next Step 5: Monitoring Your Labeling Job Label Images Bounding Box Creating a Bounding Box Labeling Job (Console) Create a Bounding Box Labeling Job (API) Provide a Template for Bounding Box Labeling Jobs Bounding Box Output Data Image Semantic Segmentation Creating a Semantic Segmentation Labeling Job (Console) Create a Semantic Segmentation Labeling Job (API) Provide a Template for Semantic Segmentation Labeling Jobs Semantic Segmentation Output Data Auto-Segmentation Tool Tool Preview Tool Availability Image Classification (Single Label) Create an Image Classification Labeling Job (Console) Create an Image Classification Labeling Job (API) Provide a Template for Image Classification Labeling Jobs Image Classification Output Data Image Classification (Multi-label) Create a Multi-Label Image Classification Labeling Job (Console) Create a Multi-Label Image Classification Labeling Job (API) Provide a Template for Multi-label Image Classification Multi-label Image Classification Output Data Image Label Verification Use Ground Truth to Label Text Named Entity Recognition Create a Named Entity Recognition Labeling Job (Console) Create a Named Entity Recognition Labeling Job (API) Provide a Template for Named Entity Recognition Labeling Jobs Named Entity Recognition Output Data Text Classification (Single Label) Create a Text Classification Labeling Job (Console) Create a Text Classification Labeling Job (API) Provide a Template for Text Classification Labeling Jobs Text Classification Output Data Text Classification (Multi-label) Create a Multi-Label Text Classification Labeling Job (Console) Create a Multi-Label Text Classification Labeling Job (API) Create a Template for Multi-label Text Classification Multi-label Text Classification Output Data Label Videos and Video Frames Video Classification Create a Video Classification Labeling Job (Console) Create a Video Classification Labeling Job (API) Provide a Template for Video Classification Video Classification Output Data Label Video Frames Video Frame Object Detection Preview the Worker UI Create a Video Frame Object Detection Labeling Job Create a Labeling Job (Console) Create a Labeling Job (API) Create Video Frame Object Detection Adjustment or Verification Labeling Job Output Data Format Video Frame Object Tracking Preview the Worker UI Create a Video Frame Object Tracking Labeling Job Create a Labeling Job (Console) Create a Labeling Job (API) Create a Video Frame Object Tracking Adjustment or Verification Labeling Job Output Data Format Video Frame Labeling Job Overview Input Data Job Completion Times Task Types Workforces Worker User Interface (UI) Label Category and Frame Attributes Label Category Attributes Frame level Attributes Worker Instructions Declining Tasks Video Frame Job Permission Requirements Add a CORS Permission Policy to S3 Bucket Worker Instructions Work on Video Frame Object Tracking Tasks Your Task Navigate the UI Bulk Edit Label and Frame Attributes Tool Guide Icons Guide Shortcuts Release, Stop and Resume, and Decline Tasks Saving Your Work and Submitting Work on Video Frame Object Detection Tasks Your Task Navigate the UI Bulk Edit Label and Frame Attributes Tool Guide UI Icon Guide Shortcuts Release, Stop and Resume, and Decline Tasks Saving Your Work and Submitting Use Ground Truth to Label 3D Point Clouds 3D Point Clouds LiDAR Sensor Fusion Label 3D Point Clouds Assistive Labeling Tools for Point Cloud Annotation Next Steps 3D Point Cloud Task types 3D Point Cloud Object Detection View the Worker Task Interface Create a 3D Point Cloud Object Detection Labeling Job Create a Labeling Job (Console) Create a Labeling Job (API) Create a 3D Point Cloud Object Detection Adjustment or Verification Labeling Job Output Data Format 3D Point Cloud Object Tracking View the Worker Task Interface Worker Tools Create a 3D Point Cloud Object Tracking Labeling Job Create a Labeling Job (API) Create a Labeling Job (Console) Create a 3D Point Cloud Object Tracking Adjustment or Verification Labeling Job Output Data Format 3D Point Cloud Semantic Segmentation View the Worker Task Interface Create a 3D Point Cloud Semantic Segmentation Labeling Job Create a Labeling Job (Console) Create a Labeling Job (API) Create a 3D Point Cloud Semantic Segmentation Adjustment or Verification Labeling Job Output Data Format 3D Point Cloud Labeling Jobs Overview Job Pre-processing Time Job Completion Times Workforces Worker User Interface (UI) Label Category Attributes Label Category Attributes Frame Attributes Worker Instructions Declining Tasks 3D Point Cloud Labeling Job Permission Requirements Add a CORS Permission Policy to S3 Bucket Worker Instructions 3D Point Cloud Semantic Segmentation Your Task Navigate the UI Icon Guide Shortcuts Release, Stop and Resume, and Decline Tasks Saving Your Work and Submitting 3D Point Cloud Object Detection Your Task Navigate the UI Icon Guide Shortcuts Release, Stop and Resume, and Decline Tasks Saving Your Work and Submitting 3D Point Cloud Object Tracking Your Task Navigate the UI Delete Cuboids Bulk Edit Label Category and Frame Attributes Icon Guide Shortcuts Release, Stop and Resume, and Decline Tasks Saving Your Work and Submitting Verify and Adjust Labels Requirements to Create Verification and Adjustment Labeling Jobs Create a Label Verification Job (Console) Create an Image Label Verification Job (Console) Create a Point Cloud or Video Frame Label Verification Job (Console) Create a Label Adjustment Job (Console) Create an Image Label Adjustment Job (Console) Create a Point Cloud or Video Frame Label Adjustment Job (Console) Start a Label Verification or Adjustment Job (API) Bounding Box and Semantic Segmentation 3D Point Cloud and Video Frame Label Verification and Adjustment Data in the Output Manifest Cautions and Considerations Color Information Requirements for Semantic Segmentation Jobs Filter Your Data Before Starting the Job Creating Custom Labeling Workflows Step 1: Setting up your workforce Next Step 2: Creating your custom worker task template Starting with a base template Developing templates locally Using External Assets Track your variables A simple sample Adding automation with Liquid Variable filters Autoescape and explicit escape escape_once skip_autoescape to_json grant_read_access End-to-end demos Next Step 3: Processing with AWS Lambda Pre-annotation and Post-annotation Lambda Function Requirements Pre-annotation Lambda Examples of Pre-annotation Lambda Functions Post-annotation Lambda Required Permissions To Use AWS Lambda With Ground Truth Grant Permission to Create and Select an AWS Lambda Function Grant IAM Execution Role Permission to Invoke AWS Lambda Functions Grant Post-Annotation Lambda Permissions to Access Annotation Create Lambda Functions for a Custom Labeling Workflow Test Pre-Annotation and Post-Annotation Lambda Functions Prerequisites Test the Pre-annotation Lambda Function Test the Post-Annotation Lambda Function Demo Template: Annotation of Images with crowd-bounding-box Starter Bounding Box custom template Your own Bounding Box custom template Your manifest file Your pre-annotation Lambda function Your post-annotation Lambda function The output of your labeling job Demo Template: Labeling Intents with crowd-classifier Starter Intent Detection custom template Your Intent Detection custom template Styling Your Elements Your pre-annotation Lambda function Your post-annotation Lambda function Your labeling job output Custom Workflows via the API Create a Labeling Job Built-in Task Types Creating Instruction Pages Short Instructions Full Instructions Add example images to your instructions Create a Labeling Job (Console) Next Steps Create a Labeling Job (API) Examples Create a Streaming Labeling Job Create Amazon SNS Input and Output Topics Create an Input Topic Create an Output Topic Add Encryption to Your Output Topic (Optional) Subscribe an Endpoint to Your Amazon SNS Output Topic Set up Amazon S3 Bucket Event Notifications Create a Manifest File (Optional) Example: Use SageMaker API To Create Streaming Labeling Job Stop a Streaming Labeling Job Create a Labeling Category Configuration File with Label Category and Frame Attributes Label Category Configuration File Schema Label and label category attribute quotas Example: Label Category Configuration Files for 3D Point Cloud Labeling Jobs Example: Label Category Configuration Files for Video Frame Labeling Jobs Creating Worker Instructions Use Input and Output Data Input Data Use an Input Manifest File Automated Data Setup Supported Data Formats Ground Truth Streaming Labeling Jobs How It Works Send Data to a Streaming Labeling Job Send Data Objects Using Amazon SNS Send Data Objects using Amazon S3 Manage Labeling Requests with an Amazon SQS Queue Receive Output Data from a Streaming Labeling Job Duplicate Message Handling Specify A Deduplication Key and ID in an Amazon SNS Message Find Deduplication Key and ID in Your Output Data Input Data Quotas Input File Size Quota Input Image Resolution Quotas Label Category Quotas 3D Point Cloud and Video Frame Labeling Job Quotas Filter and Select Data for Labeling Use the Full Dataset Choose a Random Sample Specify a Subset 3D Point Cloud Input Data Accepted Raw 3D Data Formats Compact Binary Pack Format ASCII Format Point Cloud Resolution Limits Create an Input Manifest File for a 3D Point Cloud Labeling Job Create a Point Cloud Frame Input Manifest File Include Vehicle Pose Information in Your Input Manifest Include Camera Data in Your Input Manifest Point Cloud Frame Limits Create a Point Cloud Sequence Input Manifest Parameters for Individual Point Cloud Frames Include Vehicle Pose Information in Your Input Manifest Include Camera Data in Your Input Manifest Sequence File and Point Cloud Frame Limits Understand Coordinate Systems and Sensor Fusion Coordinate System Requirements for Labeling Jobs Using Point Cloud Data in a World Coordinate System What is a World Coordinate System? Convert 3D Point Cloud Data to a WCS Sensor Fusion Extrinsic Matrix Intrinsic Matrix Image Distortion Ego Vehicle Pose Compute Orientation Quaternions and Position Ground Truth Sensor Fusion Transformations LiDAR Extrinsic Camera Calibrations: Extrinsic, Intrinsic and Distortion Camera Extrinsic Intrinsic and Distortion Video Frame Input Data Choose Video Files or Video Frames for Input Data Provide Video Frames Provide Video Files Input Data Setup Automated Video Frame Input Data Setup Provide Video Files and Extract Frames Provide Video Frames Manual Input Data Setup Create a Video Frame Input Manifest File Create a Video Frame Sequence Input Manifest Create a Video Frame Sequence File Output Data Output Directories Active Learning Directory Annotations Directory Inference Directory Manifest Directory Training Directory Confidence Score Worker Metadata Output Metadata Classification Job Output Multi-label Classification Job Output Bounding Box Job Output Named Entity Recognition Label Verification Job Output Semantic Segmentation Job Output Video Frame Object Detection Output Video Frame Object Tracking Output 3D Point Cloud Semantic Segmentation Output 3D Point Cloud Object Detection Output 3D Point Cloud Object Tracking Output Enhanced Data Labeling Control the Flow of Data Objects Sent to Workers Use MaxConcurrentTaskCount to Control the Flow of Data Objects Use Amazon SQS to Control the Flow of Data Objects to Streaming Labeling Jobs Consolidate Annotations Create Your Own Annotation Consolidation Function Assess Similarity Assess the Most Probable Label Automate Data Labeling How it Works Accuracy of Automated Labels Create an Automated Data Labeling Job (Console) Create an Automated Data Labeling Job (API) Amazon EC2 Instances Required for Automated Data Labeling Set up an active learning workflow with your own model Chaining Labeling Jobs Key Term: Label Attribute Name Start a Chained Job (Console) Job Overview Panel Start a Chained Job (API) Use a Partially Labeled Dataset Ground Truth Security and Permissions CORS Permission Requirement Assign IAM Permissions to Use Ground Truth Use IAM Managed Policies with Ground Truth Grant IAM Permission to Use the Amazon SageMaker Ground Truth Console Ground Truth Console Permissions Custom Labeling Workflow Permissions Private Workforce Permissions Vendor Workforce Permissions Create a SageMaker Execution Role for a Ground Truth Labeling Job Built-In Task Types (Non-streaming) Execution Role Requirements Built-In Task Types (Streaming) Execution Role Requirements Execution Role Requirements for Custom Task Types Automated Data Labeling Permission Requirements Encrypt Output Data and Storage Volume with AWS KMS Encrypt Output Data using KMS Encrypt Automated Data Labeling ML Compute Instance Storage Volume Output Data and Storage Volume Encryption Use Your KMS Key to Encrypt Output Data Use Your KMS Key to Encrypt Automated Data Labeling Storage Volume (API Only) Workforce Authentication and Restrictions Restrict Access to Workforce Types Monitor Labeling Job Status Send Events to CloudWatch Events Set Up a Target to Process Events Labeling Job Expiration Declining Tasks Create and Manage Workforces Using the Amazon Mechanical Turk Workforce Use Mechanical Turk with Ground Truth Use Mechanical Turk with Amazon A2I When is Mechanical Turk Not Supported? Managing Vendor Workforces Use a Private Workforce Create and Manage Amazon Cognito Workforce Create a Private Workforce (Amazon Cognito) Create a Private Workforce (Amazon SageMaker Console) Create an Amazon Cognito Workforce When Creating a Labeling Job Create an Amazon Cognito Workforce Using the Labeling Workforces Page Create a Private Workforce (Amazon Cognito Console) Manage a Private Workforce (Amazon Cognito) Manage a Workforce (Amazon SageMaker Console) Create a Work Team Using the SageMaker Console Subscriptions Add or Remove Workers Add Workers to the Workforce Add a Worker to a Work Team Disable and Remove a Worker from the Workforce Manage a Private Workforce (Amazon Cognito Console) Create Work Teams (Amazon Cognito Console) Subscriptions Add and Remove Workers (Amazon Cognito Console) Add a Worker to a Work Team Disable and Remove a Worker From a Work Team Create and Manage OIDC IdP Workforce Create a Private Workforce (OIDC IdP) Send Required and Optional Claims to Ground Truth and Amazon A2I Create an OIDC IdP Workforce Configure your OIDC IdP Validate Your OIC IdP Workforce Authentication Response Next Steps Manage a Private Workforce (OIDC IdP) Prerequisites Add work teams Add or remove IdP groups from work teams Delete a work team Manage Individual Workers Update, Delete, and Describe Your Workforce Manage Private Workforce Using the Amazon SageMaker API Find Your Workforce Name Restrict Worker Access to Tasks to Allowable IP Addresses Update OIDC Identity Provider Workforce Configuration Delete a Private Workforce Track Worker Performance Enable Tracking Examine Logs Use Log Metrics Create and manage Amazon SNS topics for your work teams Create the Amazon SNS topic Manage worker subscriptions Crowd HTML Elements Reference SageMaker Crowd HTML Elements crowd-alert Attributes dismissible type Element Hierarchy See Also crowd-badge Attributes for icon label Element Hierarchy See Also crowd-button Attributes disabled form-action href icon icon-align icon-url loading target variant Element Hierarchy See Also crowd-bounding-box Attributes header initial-value labels name src Element Hierarchy Regions full-instructions short-instructions Output boundingBoxes inputImageProperties See Also crowd-card Attributes heading image Element Hierarchy See Also crowd-checkbox Attributes checked disabled name required value Element Hierarchy Output See Also crowd-classifier Attributes categories header name Element Hierarchy Regions classification-target full-instructions short-instructions Output See Also crowd-classifier-multi-select Attributes categories header name exclusion-category Element Hierarchy Regions classification-target full-instructions short-instructions Output See Also crowd-entity-annotation Attributes header initial-value labels name text Element Hierarchy Regions full-instructions short-instructions Output entities See Also crowd-fab Attributes disabled icon label title Element Hierarchy See Also crowd-form Element Hierarchy Element Events See Also crowd-icon-button Attributes disabled icon Element Hierarchy See Also crowd-image-classifier Attributes categories header name overlay src Element Hierarchy Regions full-instructions short-instructions worker-comment header link-text placeholder Output See Also crowd-image-classifier-multi-select Attributes categories header name src exclusion-category Element Hierarchy Regions full-instructions short-instructions Output See Also crowd-input Attributes allowed-pattern auto-focus auto-validate disabled error-message label max-length min-length name placeholder required type value Element Hierarchy Output See Also crowd-instance-segmentation Attributes header labels name src initial-value Element Hierarchy Regions full-instructions short-instructions Output labeledImage instances inputImageProperties See Also crowd-instructions Attributes link-text link-type Element Hierarchy Regions detailed-instructions negative-example positive-example short-summary See Also crowd-keypoint Attributes header initial-value labels name src Element Hierarchy Regions full-instructions short-instructions Output inputImageProperties keypoints See Also crowd-line Attributes header initial-value labels label-colors name src Regions full-instructions short-instructions Element Hierarchy Output inputImageProperties lines See Also crowd-modal Attributes link-text link-type Element Hierarchy See Also crowd-polygon Attributes header labels name src initial-value Element Hierarchy Regions full-instructions short-instructions Output polygons inputImageProperties See Also crowd-polyline Attributes header initial-value labels label-colors name src Regions full-instructions short-instructions Element Hierarchy Output inputImageProperties polylines See Also crowd-radio-button Attributes checked disabled name value Element Hierarchy Output See Also crowd-radio-group Attributes Element Hierarchy Output See Also crowd-semantic-segmentation Attributes header initial-value labels name src Element Hierarchy Regions full-instructions short-instructions Output labeledImage labelMappings initialValueModified inputImageProperties See Also crowd-slider Attributes disabled editable max min name pin required secondary-progress step value Element Hierarchy See Also crowd-tab Attributes header Element Hierarchy See Also crowd-tabs Attributes Element Hierarchy See Also crowd-text-area Attributes auto-focus auto-validate char-counter disabled error-message label max-length max-rows name placeholder rows value Element Hierarchy Output See Also crowd-toast Attributes duration text Element Hierarchy See Also crowd-toggle-button Attributes checked disabled invalid name required value Element Hierarchy Output See Also Augmented AI Crowd HTML Elements crowd-textract-analyze-document Attributes header src initialValue blockTypes keys no-key-edit no-geometry-edit Element Hierarchy Regions full-instructions short-instructions Example of a Worker Template Using the crowd Element Output crowd-rekognition-detect-moderation-labels Attributes header src categories exclusion-category Element Hierarchy AWS Regions full-instructions short-instructions Example Worker Template with the crowd Element Output Prepare and Analyze Datasets Detect Pretraining Data Bias Amazon SageMaker Clarify Terms for Bias and Fairness Sample Notebooks Measure Pretraining Bias Class Imbalance (CI) Difference in Proportions of Labels (DPL) Kullback-Leibler Divergence (KL) Jensen-Shannon Divergence (JS) Lp-norm (LP) Total Variation Distance (TVD) Kolmogorov-Smirnov (KS) Conditional Demographic Disparity (CDD) Generate Reports for Bias in Pretraining Data in SageMaker Studio Prepare ML Data with Amazon SageMaker Data Wrangler Get Started with Data Wrangler Prerequisites Access Data Wrangler Update Data Wrangler Demo: Data Wrangler Titanic Dataset Walkthrough Upload Dataset to S3 and Import Data Flow Prepare and Visualize Data Exploration Drop Unused Columns Clean up Missing Values Custom Pandas: Encode Custom SQL: SELECT Columns Export Export to Data Wrangler Job Notebook Training XGBoost Classifier Shut down Data Wrangler Import Import data from Amazon S3 Import data from Athena Import data from Amazon Redshift Import data from Snowflake Administrator Guide Configure Snowflake with Data Wrangler What information needs to be provided to the Data Scientist Data Scientist Guide Private Connectivity between Data Wrangler and Snowflake via AWS PrivateLink Create a VPC Set up Snowflake AWS PrivateLink Integration Configure DNS for Snowflake Endpoints in your VPC Configure Route 53 Resolver Inbound Endpoint for your VPC SageMaker VPC Endpoints Imported Data Storage Amazon Redshift Import Storage Amazon Athena Import Storage Create and Use a Data Wrangler Flow Instances The Data Flow UI Add a Step to Your Data Flow Delete a step from your Data Flow Transform Data Transform UI Join Datasets Concatenate Datasets Custom Transforms Custom Formula Encode Categorical Ordinal Encode One-Hot Encode Featurize Text Character Statistics Vectorize Featurize Date/Time Format String Handle Outliers Robust standard deviation numeric outliers Standard Deviation Numeric Outliers Quantile Numeric Outliers Min-Max Numeric Outliers Replace Rare Handle Missing Values Fill Missing Impute Missing Add Indicator for Missing Drop Missing Manage Columns Manage Rows Manage Vectors Process Numeric Search and Edit Parse Value as Type Validate String Analyze and Visualize Histogram Scatter Plot Table Summary Quick Model Target Leakage Bias Report Create Custom Visualizations Export Export to a Data Wrangler Job Export to SageMaker Pipelines Use A Jupyter Notebook to Create a Pipeline Export to Python Code Export to the SageMaker Feature Store Use A Jupyter Notebook to Add Features to a Feature Store Export to Amazon S3 Shut Down Data Wrangler Update Data Wrangler Security and Permissions Add a Bucket Policy To Restrict Access to Datasets Imported to Data Wrangler Grant an IAM Role Permission to Use Data Wrangler Snowflake and Data Wrangler Data Encryption with KMS-CMK Amazon S3 CMK setup for Data Wrangler imported data storage Release Notes Troubleshoot Process Data Use Amazon SageMaker Processing Sample Notebooks Monitor Amazon SageMaker Processing Jobs with CloudWatch Logs and Metrics Data Processing with Apache Spark Running a Spark Processing Job Data Processing with scikit-learn Use Your Own Processing Code Run Scripts with Your Own Processing Container Build Your Own Processing Container (Advanced Scenario) How Amazon SageMaker Processing Runs Your Processing Container Image How Amazon SageMaker Processing Configures Input and Output For Your Processing Container How Amazon SageMaker Processing Provides Logs and Metrics for Your Processing Container How Amazon SageMaker Processing Configures Your Processing Container Save and Access Metadata Information About Your Processing Job Run Your Processing Container Using the SageMaker Python SDK Create, Store, and Share Features with Amazon SageMaker Feature Store How Feature Store Works Create Feature Groups Find, Discover, and Share Features Real-Time Inference for Features Stored in the Online Store Offline Store for Model Training and Batch Inference Feature Data Ingestion Get started with Amazon SageMaker Feature Store Feature Store Concepts Create Feature Groups Introduction to Feature Store Step 1: Set Up Step 2: Inspect your data Step 3: Create feature groups Step 4: Ingest data into a feature group Step 5: Clean up Step 6: Next steps Step 7: Programmers note Fraud Detection with Feature Store Step 1: Set Up Feature Store Step 2: Load Datasets and Partition Data into Feature Groups Step 3: Set Up Feature Groups Step 4: Set Up Record Identifier and Event Time Features Step 5: Load Feature Definitions Step 6: Create a Feature Group Step 7: Work with Feature Groups Describe a Feature Group List Feature Groups Put Records in a Feature Group Get Records from a Feature Group Generate Hive DDL Commands Build a Training Dataset Write and Execute an Athena Query Delete a Feature Group Adding required policies to your IAM role Step 1: Access AWS Management Console Step 2: Choose Roles Step 3: Find your role Step 4: Attach policy Use Amazon SageMaker Feature Store with Amazon SageMaker Studio Create a Feature Group in Studio View Feature Group Details in Studio Data Sources and Ingestion Stream Ingestion Data Wrangler with Feature Store Query Feature Store with Athena and AWS Glue Sample Athena Queries Cross-Account Offline Store Access Step 1: Set Up the Offline Store Access Role in Account A Step 2: Set up an Offline Store S3 Bucket in Account B Step 3: Set up an Offline Store KMS Encryption Key in Account A Step 4: Create a Feature Group in Account A Quotas, Naming Rules and Data Types Limits and Quotas Naming Rules Data Types Amazon SageMaker Feature Store Offline Store Data Format Amazon SageMaker Feature Store Notebook Examples Feature Store sample notebooks Train Models Choose an Algorithm Choose an algorithm implementation Use a built-in algorithm Use script mode in a supported framework Use a custom Docker image Problem types for the basic machine learning paradigms Supervised learning Unsupervised learning Reinforcement learning Use Amazon SageMaker Built-in Algorithms Supervised Learning Unsupervised Learning Textual Analysis Image Processing Common Information About Built-in Algorithms Docker Registry Paths and Example Code Docker Registry Paths and Example Code for US East (Ohio) (us-east-2) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) LDA (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Ray PyTorch (DLC) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) VW (algorithm) XGBoost (algorithm) Docker Registry Paths and Example Code for US East (N. Virginia) (us-east-1) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) LDA (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Ray PyTorch (DLC) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) VW (algorithm) XGBoost (algorithm) Docker Registry Paths and Example Code for US West (N. California) (us-west-1) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) LDA (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Ray PyTorch (DLC) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) VW (algorithm) XGBoost (algorithm) Docker Registry Paths and Example Code for US West (Oregon) (us-west-2) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) LDA (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Ray PyTorch (DLC) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) VW (algorithm) XGBoost (algorithm) Docker Registry Paths and Example Code for Africa (Cape Town) (af-south-1) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) XGBoost (algorithm) Docker Registry Paths and Example Code for Asia Pacific (Hong Kong) (ap-east-1) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) XGBoost (algorithm) Docker Registry Paths and Example Code for Asia Pacific (Mumbai) (ap-south-1) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) LDA (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Ray PyTorch (DLC) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) VW (algorithm) XGBoost (algorithm) Docker Registry Paths and Example Code for Asia Pacific (Seoul) (ap-northeast-2) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) LDA (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Ray PyTorch (DLC) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) VW (algorithm) XGBoost (algorithm) Docker Registry Paths and Example Code for Asia Pacific (Singapore) (ap-southeast-1) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) LDA (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Ray PyTorch (DLC) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) VW (algorithm) XGBoost (algorithm) Docker Registry Paths and Example Code for Asia Pacific (Sydney) (ap-southeast-2) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) LDA (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Ray PyTorch (DLC) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) VW (algorithm) XGBoost (algorithm) Docker Registry Paths and Example Code for Asia Pacific (Tokyo) (ap-northeast-1) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) LDA (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Ray PyTorch (DLC) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) VW (algorithm) XGBoost (algorithm) Docker Registry Paths and Example Code for Canada (Central) (ca-central-1) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) LDA (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Ray PyTorch (DLC) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) VW (algorithm) XGBoost (algorithm) Docker Registry Paths and Example Code for China (Beijing) (cn-north-1) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) XGBoost (algorithm) Docker Registry Paths and Example Code for China (Ningxia) (cn-northwest-1) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) XGBoost (algorithm) Docker Registry Paths and Example Code for Europe (Frankfurt) (eu-central-1) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) LDA (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Ray PyTorch (DLC) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) VW (algorithm) XGBoost (algorithm) Docker Registry Paths and Example Code for Europe (Ireland) (eu-west-1) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) LDA (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Ray PyTorch (DLC) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) VW (algorithm) XGBoost (algorithm) Docker Registry Paths and Example Code for Europe (London) (eu-west-2) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) LDA (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Ray PyTorch (DLC) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) VW (algorithm) XGBoost (algorithm) Docker Registry Paths and Example Code for Europe (Paris) (eu-west-3) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) XGBoost (algorithm) Docker Registry Paths and Example Code for Europe (Stockholm) (eu-north-1) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) XGBoost (algorithm) Docker Registry Paths and Example Code for Europe (Milan) (eu-south-1) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) XGBoost (algorithm) Docker Registry Paths and Example Code for Middle East (Bahrain) (me-south-1) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) XGBoost (algorithm) Docker Registry Paths and Example Code for South America (São Paulo) (sa-east-1) BlazingText (algorithm) Chainer (DLC) Clarify (algorithm) Data Wrangler (algorithm) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) Model Monitor (algorithm) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) XGBoost (algorithm) Docker Registry Paths and Example Code for AWS GovCloud (US-West) (us-gov-west-1) BlazingText (algorithm) Chainer (DLC) Debugger (algorithm) DeepAR Forecasting (algorithm) Factorization Machines (algorithm) Hugging Face (algorithm) IP Insights (algorithm) Image classification (algorithm) Inferentia MXNet (DLC) Inferentia PyTorch (DLC) K-Means (algorithm) KNN (algorithm) LDA (algorithm) Linear Learner (algorithm) MXNet (DLC) MXNet Coach (DLC) NTM (algorithm) Neo Image Classification (algorithm) Neo MXNet (DLC) Neo PyTorch (DLC) Neo Tensorflow (DLC) Neo XGBoost (algorithm) Object Detection (algorithm) Object2Vec (algorithm) PCA (algorithm) PyTorch (DLC) Random Cut Forest (algorithm) Scikit-learn (algorithm) Semantic Segmentation (algorithm) Seq2Seq (algorithm) Spark (algorithm) SparkML Serving (algorithm) Tensorflow (DLC) Tensorflow Coach (DLC) Tensorflow Inferentia (DLC) Tensorflow Ray (DLC) XGBoost (algorithm) Common Data Formats for Built-in Algorithms Common Data Formats for Training Content Types Supported by Built-In Algorithms Using Pipe Mode Using CSV Format Using RecordIO Format Trained Model Deserialization Common Data Formats for Inference Convert Data for Inference Request Serialization Convert Data for Inference Response Deserialization Common Request Formats for All Algorithms JSON Request Format JSONLINES Request Format CSV Request Format RECORDIO Request Format Use Batch Transform with Built-in Algorithms Instance Types for Built-in Algorithms Logs for Built-in Algorithms Common Errors BlazingText algorithm Input/Output Interface for the BlazingText Algorithm Training and Validation Data Format Training and Validation Data Format for the Word2Vec Algorithm Training and Validation Data Format for the Text Classification Algorithm Train with File Mode Train with Augmented Manifest Text Format Model Artifacts and Inference Model Artifacts for the Word2Vec Algorithm Sample JSON Request Model Artifacts for the Text Classification Algorithm Sample JSON Request EC2 Instance Recommendation for the BlazingText Algorithm BlazingText Sample Notebooks BlazingText Hyperparameters Word2Vec Hyperparameters Text Classification Hyperparameters Tune a BlazingText Model Metrics Computed by the BlazingText Algorithm Tunable BlazingText Hyperparameters Tunable Hyperparameters for the Word2Vec Algorithm Tunable Hyperparameters for the Text Classification Algorithm DeepAR Forecasting Algorithm Input/Output Interface for the DeepAR Algorithm Best Practices for Using the DeepAR Algorithm EC2 Instance Recommendations for the DeepAR Algorithm DeepAR Sample Notebooks How the DeepAR Algorithm Works How Feature Time Series Work in the DeepAR Algorithm DeepAR Hyperparameters Tune a DeepAR Model Metrics Computed by the DeepAR Algorithm Tunable Hyperparameters for the DeepAR Algorithm DeepAR Inference Formats DeepAR JSON Request Formats DeepAR JSON Response Formats Batch Transform with the DeepAR Algorithm Factorization Machines Algorithm Input/Output Interface for the Factorization Machines Algorithm EC2 Instance Recommendation for the Factorization Machines Algorithm Factorization Machines Sample Notebooks How Factorization Machines Work Factorization Machines Hyperparameters Tune a Factorization Machines Model Metrics Computed by the Factorization Machines Algorithm Tunable Factorization Machines Hyperparameters Factorization Machines Response Formats JSON Response Format JSONLINES Response Format RECORDIO Response Format Image Classification Algorithm Input/Output Interface for the Image Classification Algorithm Train with RecordIO Format Train with Image Format Train with Augmented Manifest Image Format Incremental Training Inference with the Image Classification Algorithm EC2 Instance Recommendation for the Image Classification Algorithm Image Classification Sample Notebooks How Image Classification Works Image Classification Hyperparameters Tune an Image Classification Model Metrics Computed by the Image Classification Algorithm Tunable Image Classification Hyperparameters IP Insights Input/Output Interface for the IP Insights Algorithm EC2 Instance Recommendation for the IP Insights Algorithm GPU Instances for the IP Insights Algorithm CPU Instances for the IP Insights Algorithm IP Insights Sample Notebooks How IP Insights Works IP Insights Hyperparameters Tune an IP Insights Model Metrics Computed by the IP Insights Algorithm Tunable IP Insights Hyperparameters IP Insights Data Formats IP Insights Training Data Formats IP Insights Training Data Input Formats INPUT: CSV IP Insights Inference Data Formats IP Insights Input Request Formats INPUT: CSV Format INPUT: JSON Format INPUT: JSONLINES Format IP Insights Output Response Formats OUTPUT: JSON Response Format OUTPUT: JSONLINES Response Format K-Means Algorithm Input/Output Interface for the K-Means Algorithm EC2 Instance Recommendation for the K-Means Algorithm K-Means Sample Notebooks How K-Means Clustering Works Step 1: Determine the Initial Cluster Centers Step 2: Iterate over the Training Dataset and Calculate Cluster Centers Step 3: Reduce the Clusters from K to k K-Means Hyperparameters Tune a K-Means Model Metrics Computed by the K-Means Algorithm Tunable K-Means Hyperparameters K-Means Response Formats JSON Response Format JSONLINES Response Format RECORDIO Response Format CSV Response Format K-Nearest Neighbors (k-NN) Algorithm Input/Output Interface for the k-NN Algorithm k-NN Sample Notebooks How the k-NN Algorithm Works Step 1: Sample Step 2: Perform Dimension Reduction Step 3: Build an Index Serialize the Model EC2 Instance Recommendation for the k-NN Algorithm Instance Recommendation for Training with the k-NN Algorithm Instance Recommendation for Inference with the k-NN Algorithm k-NN Hyperparameters Tune a k-NN Model Metrics Computed by the k-NN Algorithm Tunable k-NN Hyperparameters Data Formats for k-NN Training Input CSV Data Format RECORDIO Data Format k-NN Request and Response Formats INPUT: CSV Request Format INPUT: JSON Request Format INPUT: JSONLINES Request Format INPUT: RECORDIO Request Format OUTPUT: JSON Response Format OUTPUT: JSONLINES Response Format OUTPUT: VERBOSE JSON Response Format OUTPUT: RECORDIO-PROTOBUF Response Format OUTPUT: VERBOSE RECORDIO-PROTOBUF Response Format SAMPLE OUTPUT for the k-NN Algorithm Latent Dirichlet Allocation (LDA) Algorithm Choosing between Latent Dirichlet Allocation (LDA) and Neural Topic Model (NTM) Input/Output Interface for the LDA Algorithm EC2 Instance Recommendation for the LDA Algorithm LDA Sample Notebooks How LDA Works LDA Hyperparameters Tune an LDA Model Metrics Computed by the LDA Algorithm Tunable LDA Hyperparameters Linear Learner Algorithm Input/Output interface for the linear learner algorithm EC2 instance recommendation for the linear learner algorithm Linear learner sample notebooks How linear learner works Step 1: Preprocess Step 2: Train Step 3: Validate and set the threshold Step 4: Deploy a trained linear model Linear learner hyperparameters Tune a linear learner model Metrics computed by the linear learner algorithm Tuning linear learner hyperparameters Linear learner response formats JSON response formats JSONLINES response formats RECORDIO response formats Neural Topic Model (NTM) Algorithm Input/Output Interface for the NTM Algorithm EC2 Instance Recommendation for the NTM Algorithm NTM Sample Notebooks NTM Hyperparameters Tune an NTM Model Metrics Computed by the NTM Algorithm Tunable NTM Hyperparameters NTM Response Formats JSON Response Format JSONLINES Response Format RECORDIO Response Format Object2Vec Algorithm I/O Interface for the Object2Vec Algorithm EC2 Instance Recommendation for the Object2Vec Algorithm Instance Recommendation for Training Instance Recommendation for Inference Object2Vec Sample Notebooks How Object2Vec Works Step 1: Process Data Step 2: Train a Model Step 3: Produce Inferences Object2Vec Hyperparameters Tune an Object2Vec Model Metrics Computed by the Object2Vec Algorithm Regressor Metrics Computed by the Object2Vec Algorithm Classification Metrics Computed by the Object2Vec Algorithm Tunable Object2Vec Hyperparameters Data Formats for Object2Vec Training Input: JSON Lines Request Format Data Formats for Object2Vec Inference GPU optimization: Classification or Regression Input: Classification or Regression Request Format Output: Classification or Regression Response Format Encoder Embeddings for Object2Vec GPU optimization: Encoder Embeddings Input: Encoder Embeddings Output: Encoder Embeddings Object Detection Algorithm Input/Output Interface for the Object Detection Algorithm Train with the RecordIO Format Train with the Image Format Train with Augmented Manifest Image Format Incremental Training EC2 Instance Recommendation for the Object Detection Algorithm Object Detection Sample Notebooks How Object Detection Works Object Detection Hyperparameters Tune an Object Detection Model Metrics Computed by the Object Detection Algorithm Tunable Object Detection Hyperparameters Object Detection Request and Response Formats Request Format Response Formats OUTPUT: JSON Response Format Principal Component Analysis (PCA) Algorithm Input/Output Interface for the PCA Algorithm EC2 Instance Recommendation for the PCA Algorithm PCA Sample Notebooks How PCA Works Mode 1: Regular Mode 2: Randomized PCA Hyperparameters PCA Response Formats JSON Response Format JSONLINES Response Format RECORDIO Response Format Random Cut Forest (RCF) Algorithm Input/Output Interface for the RCF Algorithm Instance Recommendations for the RCF Algorithm RCF Sample Notebooks How RCF Works Sample Data Randomly Train a RCF Model and Produce Inferences Choose Hyperparameters References RCF Hyperparameters Tune an RCF Model Metrics Computed by the RCF Algorithm Tunable RCF Hyperparameters RCF Response Formats JSON Response Format JSONLINES Response Format RECORDIO Response Format Semantic Segmentation Algorithm Semantic Segmentation Sample Notebooks Input/Output Interface for the Semantic Segmentation Algorithm How Training Works Training with the Augmented Manifest Format Incremental Training Produce Inferences EC2 Instance Recommendation for the Semantic Segmentation Algorithm Semantic Segmentation Hyperparameters Tuning a Semantic Segmentation Model Metrics Computed by the Semantic Segmentation Algorithm Tunable Semantic Segmentation Hyperparameters Sequence-to-Sequence Algorithm Input/Output Interface for the Sequence-to-Sequence Algorithm EC2 Instance Recommendation for the Sequence-to-Sequence Algorithm Sequence-to-Sequence Sample Notebooks How Sequence-to-Sequence Works Sequence-to-Sequence Hyperparameters Tune a Sequence-to-Sequence Model Metrics Computed by the Sequence-to-Sequence Algorithm Tunable Sequence-to-Sequence Hyperparameters XGBoost Algorithm Supported versions How to Use SageMaker XGBoost Input/Output Interface for the XGBoost Algorithm EC2 Instance Recommendation for the XGBoost Algorithm XGBoost Sample Notebooks How XGBoost Works XGBoost Hyperparameters Tune an XGBoost Model Evaluation Metrics Computed by the XGBoost Algorithm Tunable XGBoost Hyperparameters Deprecated Versions of XGBoost and their Upgrades Upgrade XGBoost Version 0.90 to Version 1.2 Upgrade SageMaker Python SDK Version 1.x to Version 2.x Change the image tag to 1.2-2 Change Docker Image for Boto3 Update Hyperparameters and Learning Objectives XGBoost Version 0.72 Input/Output Interface for the XGBoost Release 0.72 EC2 Instance Recommendation for the XGBoost Release 0.72 XGBoost Release 0.72 Sample Notebooks XGBoost Release 0.72 Hyperparameters Tune an XGBoost Release 0.72 Model Metrics Computed by the XGBoost Release 0.72 Algorithm Tunable XGBoost Release 0.72 Hyperparameters Use Reinforcement Learning with Amazon SageMaker What are the differences between reinforcement, supervised, and unsupervised learning paradigms? Why is Reinforcement Learning Important? Markov Decision Process (MDP) Key Features of Amazon SageMaker RL Reinforcement Learning Sample Notebooks Sample RL Workflow Using Amazon SageMaker RL RL Environments in Amazon SageMaker Use OpenAI Gym Interface for Environments in SageMaker RL Use Open-Source Environments Use Commercial Environments Distributed Training with Amazon SageMaker RL Hyperparameter Tuning with Amazon SageMaker RL Manage Machine Learning with Amazon SageMaker Experiments SageMaker Experiments Features Organize Experiments Track Experiments Compare and Evaluate Experiments Amazon SageMaker Autopilot Create an Amazon SageMaker Experiment View and Compare Amazon SageMaker Experiments, Trials, and Trial Components View Experiments, Trials, and Trial Components Compare Experiments, Trials, and Trial Components Track and Compare Tutorial Open the Notebook in Studio Install the Experiments SDK and Import Modules Transform and Track the Input Data Create and Track an Experiment Compare and Analyze Trials Search Experiments Using Amazon SageMaker Studio Search Experiments, Trials, and Trial Components Search the SageMaker Studio Leaderboard Search by Tag Clean Up Amazon SageMaker Experiment Resources Clean Up Using the Experiments SDK Clean Up Using the Python SDK (Boto3) Search Using the Amazon SageMaker Console and API Sample Notebooks for Managing ML Experiments Organize, Find, and Evaluate Training Jobs (Console) Use Tags to Track Training Jobs (Console) Find Training Jobs (Console) Evaluate Models (Console) Find and Evaluate Training Jobs (API) Find Training Jobs (API) Evaluate Models (API) Get Suggestions for a Search (API) Verify the Datasets Used by Your Training Jobs Trace Model Lineage Trace Model Lineage (Console) Trace Model Lineage (API) Amazon SageMaker Debugger Amazon SageMaker Debugger Features Supported Frameworks and Algorithms Use Debugger with Custom Training Containers Debugger Open-Source GitHub Repositories Amazon SageMaker Debugger Architecture Get Started with Debugger Tutorials Debugger Tutorial Videos Analyze, Detect, and Get Alerted on Problems with Training Runs Using Amazon SageMaker Debugger Debug Models with Amazon SageMaker Debugger in Studio Deep Dive on Amazon SageMaker Debugger and SageMaker Model Monitor Debugger Example Notebooks Debugger Example Notebooks for Profiling Training Jobs Debugger Example Notebooks for Analyzing Model Parameters Debugger Advanced Demos and Visualization Train and Tune Your Models with Amazon SageMaker Experiments and Debugger Using SageMaker Debugger to Monitor a Convolutional Autoencoder Model Training Using SageMaker Debugger to Monitor Attentions in BERT Model Training Using SageMaker Debugger to Visualize Class Activation Maps in Convolutional Neural Networks (CNNs) Configure Debugger Using Amazon SageMaker Python SDK Construct a SageMaker Estimator with Debugger Configure Debugger Monitoring Hardware System Resource Utilization Configure Debugger Framework Profiling Start a Training Job with the Default System Monitoring and Framework Profiling Start a Training Job with the Default System Monitoring and Customized Framework Profiling for Target Steps or a Target Time Range Start a Training Job with the Default System Monitoring and Customized Framework Profiling with Different Profiling Options Updating Debugger System Monitoring and Framework Profiling Configuration while a Training Job is Running Configure Debugger Hook to Save Tensors Configure Debugger Tensor Collections Using the CollectionConfig API Operation Configure Debugger Hook to Save Tensors Example Notebooks and Code Samples to Configure Debugger Hook Tensor Visualization Example Notebooks Save Tensors Using Debugger Built-in Collections Save Tensors Using Debugger Modified Built-in Collections Save Tensors Using Debugger Custom Collections Configure Debugger Built-in Rules Use Debugger Built-in Rules with the Default Parameter Settings Use Debugger Built-in Rules with Custom Parameter Values Example Notebooks and Code Samples to Configure Debugger Rules Debugger Built-in Rules Example Notebooks Debugger Built-in Rules Example Code Use Debugger Built-in Rules with Parameter Modifications Turn Off Debugger Useful SageMaker Estimator Classmethods for Debugger Configure Debugger Using Amazon SageMaker API JSON (AWS CLI) To configure a Debugger rule for debugging model parameters To configure a Debugger built-in rule for profiling system and framework metrics Update Debugger Profiling Configuration Using the UpdateTrainingJob API Operation Add Debugger Custom Rule Configuration to the CreateTrainingJob API Operation AWS Boto3 To configure a Debugger rule for debugging model parameters To configure a Debugger built-in rule for profiling system and framework metrics Update Debugger Profiling Configuration Using the UpdateTrainingJob API Operation Add Debugger Custom Rule Configuration to the CreateTrainingJob API Operation List of Debugger Built-in Rules Debugger ProfilerRule Debugger Rule ProfilerReport BatchSize CPUBottleneck GPUMemoryIncrease IOBottleneck LoadBalancing LowGPUUtilization OverallSystemUsage MaxInitializationTime OverallFrameworkMetrics StepOutlier CreateXgboostReport DeadRelu ExplodingTensor PoorWeightInitialization SaturatedActivation VanishingGradient WeightUpdateRatio AllZero ClassImbalance LossNotDecreasing Overfit Overtraining SimilarAcrossRuns StalledTrainingRule TensorVariance UnchangedTensor CheckInputImages NLPSequenceRatio Confusion FeatureImportanceOverweight TreeDepth Create Debugger Custom Rules for Training Job Analysis Prerequisites for Creating Debugger Custom Rules Use the Debugger Client Library smdebug to Create a Custom Rule Python Script Use the Debugger APIs to Run Your Own Custom Rules Use Debugger with Custom Training Containers Prepare to Build a Custom Training Container Register Debugger Hook to Your Training Script Create and Configure a Dockerfile Build and Push the Custom Training Container to Amazon ECR Run and Debug Training Jobs Using the Custom Training Container Action on Amazon SageMaker Debugger Rules Debugger Built-in Actions for Rules Step 1: Set Up Amazon SNS, Create an SMDebugRules Topic, and Subscribe to the Topic Step 2: Set Up Your IAM Role to Attach Required Policies Step 3: Configure Debugger Rules with the Built-in Actions Considerations for Using the Debugger Built-in Actions Create Actions on Rules Using Amazon CloudWatch and AWS Lambda CloudWatch Logs for Debugger Rules and Training Jobs Set Up Debugger for Automated Training Job Termination Using CloudWatch and Lambda Step 1: Create a Lambda Function Step 2: Configure the Lambda function Step 3: Create a CloudWatch Events Rule and Link to the Lambda Function for Debugger Run Example Notebooks to Test Automated Training Job Termination Disable the CloudWatch Events Rule to Stop Using the Automated Training Job Termination SageMaker Debugger on Studio Open Amazon SageMaker Debugger Insights Dashboard SageMaker Debugger Insights Dashboard Controller SageMaker Debugger Insights Controller UI Enable and Configure Debugger Profiling for Detailed Insights SageMaker Debugger Insights Dashboard Walkthrough Debugger Insights – Overview Training job summary Resource utilization summary Resource intensive operations Insights Debugger Insights – Nodes Shut Down the SageMaker Debugger Insights Instance SageMaker Debugger on Studio Experiments Visualize Tensors Using SageMaker Debugger and Studio Loss Curves While Training Is in Progress Analyzing Training Jobs: Comparing Loss Curves Across Multiple Jobs Rules Triggering and Logs from Jobs SageMaker Debugger Interactive Reports SageMaker Debugger Profiling Report Download a Debugger Profiling Report Debugger Profiling Report Walkthrough Training Job Summary System Usage Statistics Framework metrics summary Overview: CPU Operators Overview: GPU Operators Rules Summary Analyzing the Training Loop – Step Durations GPU Utilization Analysis Batch Size CPU Bottlenecks I/O Bottlenecks LoadBalancing in Multi-GPU Training GPU Memory Analysis SageMaker Debugger XGBoost Training Report Construct a SageMaker XGBoost Estimator with the Debugger XGBoost Report Rule Download the Debugger XGBoost Training Report Debugger XGBoost Training Report Walkthrough Distribution of True Labels of the Dataset Loss versus Step Graph Feature Importance Confusion Matrix Evaluation of the Confusion Matrix Accuracy Rate of Each Diagonal Element Over Iteration Receiver Operating Characteristic Curve Distribution of Residuals at the Last Saved Step Absolute Validation Error per Label Bin Over Iteration Analyze Data Using the SMDebug Client Library Access the Monitoring and Profiling Data Plot the System Metrics and Framework Metrics Data Access the Profiling Data Using the Pandas Data Parsing Tool Access the Python Profiling Stats Data Merge Timelines of Different Profiling Trace Files Profiling Data Loader Visualize Amazon SageMaker Debugger Output Tensors in TensorBoard Best Practices for Amazon SageMaker Debugger Choose a Machine Learning Framework Use Studio Debugger Insights Dashboard Download Debugger Reports and Gain More Insights Capture Data from Your Training Job and Save Data to Amazon S3 Analyze the Data with a Fleet of Debugger Built-in Rules Take Actions Based on the Built-in Rule Status Dive Deep into the Data Using the SMDebug Client Library Monitoring System Utilization and Detect Bottlenecks Profiling Framework Operations Debugging Model Parameters Amazon SageMaker Debugger Advanced Topics and Reference Documentation Amazon SageMaker Debugger API Operations Use Debugger Docker Images for Built-in or Custom Rules Amazon SageMaker Debugger Registry URLs for Built-in Rule Evaluators Amazon SageMaker Debugger Registry URLs for Custom Rule Evaluators Amazon SageMaker Debugger Exceptions Considerations for Amazon SageMaker Debugger Considerations for Distributed Training Considerations for Monitoring System Bottlenecks and Profiling Framework Operations Considerations for Debugging Model Output Tensors Amazon SageMaker Debugger Usage Statistics Debugger Profiling Report Usage (Recommended) Option 1: Opt Out before Running a Training Job Option 2: Opt Out after a Training Job Has Completed Perform Automatic Model Tuning with SageMaker How Hyperparameter Tuning Works Random Search Bayesian Search Define Metrics Define Hyperparameter Ranges Hyperparameter Scaling Tune Multiple Algorithms with Hyperparameter Optimization to Find the Best Model Get Started Create a Hyperparameter Optimization Tuning Job for One or More Algorithms (Console) Define job settings Create Training Job Definitions Configure algorithm and parameters Define Data Input and Output Configure Training Job Resources Add or Clone a Training Job Configure Tuning Job Resources Review and Create HPO Tuning Job Manage Hyperparameter Tuning and Training Jobs Example: Hyperparameter Tuning Job Prerequisites Create a Notebook Next Step Get the Amazon SageMaker Boto 3 Client Next Step Get the SageMaker Execution Role Next Step Specify a S3 Bucket to Upload Training Datasets and Store Output Data Next Step Download, Prepare, and Upload Training Data Download and Explore the Training Dataset Prepare and Upload Data Next Step Configure and Launch a Hyperparameter Tuning Job Specify the Hyperparameter Tuning Job Settings Configure the Training Jobs Name and Launch the Hyperparameter Tuning Job Next Step Monitor the Progress of a Hyperparameter Tuning Job View the Status of the Hyperparameter Tuning Job View the Status of the Training Jobs View the Best Training Job Next Step Clean up Stop Training Jobs Early How Early Stopping Works Algorithms That Support Early Stopping Run a Warm Start Hyperparameter Tuning Job Types of Warm Start Tuning Jobs Warm Start Tuning Restrictions Warm Start Tuning Sample Notebook Create a Warm Start Tuning Job Create a Warm Start Tuning Job ( Low-level SageMaker API for Python (Boto 3)) Create a Warm Start Tuning Job (SageMaker Python SDK) Resource Limits for Automatic Model Tuning Best Practices for Hyperparameter Tuning Choosing the Number of Hyperparameters Choosing Hyperparameter Ranges Using Logarithmic Scales for Hyperparameters Choosing the Best Number of Concurrent Training Jobs Running Training Jobs on Multiple Instances Amazon SageMaker Distributed Training Libraries Get Started with Distributed Training Basic Distributed Training Concepts Advanced Concepts Strategies Train with Data Parallel and Model Parallel Optimize Distributed Training Batch Size Mini-Batch Size Scenarios Scaling from a Single GPU to Many GPUs Scaling from a Single Instance to Multiple Instances Availability Zones and Network Backplane Optimized GPU, Network, and Storage Custom Training Scripts SageMaker Built-In Distributed Training Features SageMaker's Distributed Data Parallel Library Introduction to SageMaker's Distributed Data Parallel Library Why Use SageMaker Distributed Data Parallel Library? Training Benchmarks Optimal Bandwidth Use with Balanced Fusion Buffer Optimal GPU Usage with Efficient AllReduce Overlapping with a Backward Pass SageMaker Distributed Data Parallel Architecture Modify Your Training Script Using the SageMaker Data Parallel Library Script Modification Overview Modify a TensorFlow Training Script Modify a PyTorch Training Script Run a SageMaker Distributed Data Parallel Training Job Use SageMaker's Distributed Data Parallel Library Use the Data Parallel Library with SageMaker's Python SDK TensorFlow Estimator PyTorch Estimator SageMaker distributed data parallel Configuration Tips and Pitfalls Data Preprocessing Single vs Multiple Nodes Debug Scaling Efficiency with Debugger Batch Size Custom MPI Options Data Parallel Library FAQ Data Parallel Troubleshooting Considerations for Using SageMaker Distributed Data Parallel with SageMaker Debugger and Checkpoints An Unexpected Prefix (model for example) Is Attached to state_dict keys (model parameters) from a PyTorch Distributed Training Job SageMaker's Distributed Model Parallel Introduction to Model Parallelism What is Model Parallelism? Important Considerations when Using Model Parallelism Core Features of SageMaker Distributed Model Parallel Automated Model Splitting How It Works Automated Model Splitting with PyTorch Automated Model Splitting with TensorFlow Comparison of Automated Model Splitting Between Frameworks Manual Model Splitting Pipeline Execution Schedule Interleaved Pipeline Simple Pipeline Pipelining Execution in Specific Frameworks Pipeline Execution with TensorFlow Pipeline Execution with PyTorch Modify Your Training Script Using SageMaker's Distributed Model Parallel Library Modify a TensorFlow Training Script Unsupported Framework Features TensorFlow TensorFlow with Horovod Manual partitioning with TensorFlow Modify a PyTorch Training Script Important Considerations Unsupported Framework Features PyTorch Manual Partitioning with PyTorch Run a SageMaker Distributed Model Parallel Training Job Launch a Training Job with the SageMaker Python SDK Extend or Adapt A Docker Container that Contains SageMaker's Distributed Model Parallel Library SageMaker distributed model parallel Configuration Tips and Pitfalls Model Parallel Troubleshooting Considerations for Using SageMaker Debugger with SageMaker Distributed Model Parallel Saving Checkpoints Convergence Using Model Parallel and TensorFlow Amazon SageMaker Distributed Training Notebook Examples Blogs and Case Studies PyTorch Examples TensorFlow Examples HuggingFace Examples How to Access or Download the SageMaker Distributed Training Notebook Examples Option 1: Use a SageMaker notebook instance Option 2: Clone the SageMaker example repository to SageMaker Studio or notebook instance Detect Posttraining Data and Model Bias Sample Notebooks Measure Posttraining Data and Model Bias Difference in Positive Proportions in Predicted Labels (DPPL) Disparate Impact (DI) Difference in Conditional Acceptance (DCAcc) Difference in Conditional Rejection (DCR) Recall Difference (RD) Difference in Acceptance Rates (DAR) Difference in Rejection Rates (DRR) Accuracy Difference (AD) Treatment Equality (TE) Conditional Demographic Disparity in Predicted Labels (CDDPL) Counterfactual Fliptest (FT) Configure an Amazon SageMaker Clarify Processing Jobs for Fairness and Explainability Prerequisites Getting Started with a SageMaker Clarify Container How It Works Configure a Processing Job Container's Input and Output Parameters Configure the Analysis Example Analysis Configuration JSON File for a CSV Dataset Example Analysis Configuration JSON File for a JSONLines Dataset Run SageMaker Clarify Processing Jobs for Bias Analysis and Explainability Compute Resources Required for SageMaker Clarify Processing Jobs Run the Clarify Processing Job Run the Clarify Processing Job with Spark Get the Analysis Results Troubleshoot SageMaker Clarify Processing Jobs Processing job fails to finish Processing job finishes without results and you get a CloudWatch warning message Error message for invalid analysis configuration Bias metric computation fails for several or all metrics Mismatch between analysis config and dataset/model input/output Model returns 500 Internal Server Error or container falls back to per-record predictions due to model error Execution role is invalid Failed to download data Could not connect to SageMaker Model Explainability Feature Attributions that Use Shapley Values SHAP Baselines for Explainability Create Feature Attribute Baselines and Explainability Reports Incremental Training in Amazon SageMaker Perform Incremental Training (Console) Perform Incremental Training (API) Managed Spot Training in Amazon SageMaker Using Managed Spot Training Managed Spot Training Lifecycle Use Checkpoints in Amazon SageMaker Checkpoints for Frameworks and Algorithms in SageMaker Enable Checkpointing Browse Checkpoint Files Resume Training From a Checkpoint Considerations for Checkpointing Provide Dataset Metadata to Training Jobs with an Augmented Manifest File Augmented Manifest File Format Stream Augmented Manifest File Data Use an Augmented Manifest File (Console) Use an Augmented Manifest File (API) Monitor and Analyze Training Jobs Using Metrics Training Metrics Sample Notebooks Defining Training Metrics Defining Regular Expressions for Metrics Defining Training Metrics (Low-level SageMaker API) Defining Training Metrics (SageMaker Python SDK) Define Training Metrics (Console) Monitoring Training Job Metrics ( Console) Monitoring Training Job Metrics (SageMaker Console) Example: Viewing a Training and Validation Curve Deploy Models for Inference Prerequisites What do you want to do? Manage Model Deployments Deploy Your Own Inference Code Guide to SageMaker Use Amazon SageMaker Elastic Inference (EI) How EI Works Choose an EI Accelerator Type Use EI in a SageMaker Notebook Instance Use EI on a Hosted Endpoint Frameworks that Support EI Use EI with SageMaker Built-in Algorithms EI Sample Notebooks Set Up to Use EI Set Up Required Permissions Use a Custom VPC to Connect to EI Set up Security Groups to Connect to EI Set up a VPC Interface Endpoint to Connect to EI Attach EI to a Notebook Instance Set Up to Use EI Use EI in Local Mode in SageMaker Use EI in Local Mode with SageMaker TensorFlow Estimators and Models Use EI in Local Mode with SageMaker Apache MXNet Estimators and Models Use EI in Local Mode with SageMaker PyTorch Estimators and Models Use EI on Amazon SageMaker Hosted Endpoints Use EI with a SageMaker TensorFlow Container Use an Estimator Object Use a Model Object Use EI with a SageMaker MXNet Container Use an Estimator Object Use a Model Object Use EI with a SageMaker PyTorch Container Use an Estimator Object Use a Model Object Use EI with Your Own Container Import the EI Version of TensorFlow, MXNet, or PyTorch into Your Docker Container Create an EI Endpoint with AWS SDK for Python (Boto 3) Create an Endpoint Configuration Create an Endpoint Asynchronous Inference How It Works How Do I Get Started? Create, Invoke, and Update an Asynchronous Endpoint Prerequisites Create an Asynchronous Inference Endpoint Create a Model Create an Endpoint Configuration Create Endpoint Invoke an Asynchronous Endpoint Update an Asynchronous Endpoint Delete an Asynchronous Endpoint Monitor Asynchronous Endpoint Monitoring with CloudWatch Common Endpoint Metrics Asynchronous Inference Endpoint Metrics Logs Check Prediction Results Amazon SNS Topics Check Your S3 Bucket Autoscale an Asynchronous Endpoint Define a Scaling Policy Define a Scaling Policy that Scales to 0 Use Batch Transform Use Batch Transform to Get Inferences from Large Datasets Speed up a Batch Transform Job Use Batch Transform to Test Production Variants Batch Transform Errors Batch Transform Sample Notebooks Associate Prediction Results with Input Records Workflow for Associating Inferences with Input Records Use Data Processing in Batch Transform Jobs Supported JSONPath Operators Batch Transform Examples Example: Output Only Inferences Example: Output Input Data and Inferences Example: Output an ID Column with Results and Exclude the ID Column from the Input (CSV) Example: Output an ID Attribute with Results and Exclude the ID Attribute from the Input (JSON) Host Multiple Models with Multi-Model Endpoints Supported Algorithms and Frameworks Sample Notebooks for Multi-Model Endpoints How Multi-Model Endpoints Work Setting SageMaker Multi-Model Endpoint Model Caching Behavior Instance Recommendations for Multi-Model Endpoint Deployments Create a Multi-Model Endpoint Create a Multi-Model Endpoint (Console) Create a Multi-Model Endpoint (AWS SDK for Python (Boto)) Invoke a Multi-Model Endpoint Retry Requests on ModelNotReadyException Errors Add or Remove Models Build Your Own Container with Multi Model Server Use the SageMaker Inference Toolkit Contract for Custom Containers to Serve Multiple Models Load Model API List Model API Get Model API Unload Model API Invoke Model API Multi-Model Endpoint Security CloudWatch Metrics for Multi-Model Endpoint Deployments Deploy multi-container endpoints Create a multi-container endpoint (Boto 3) Update a multi-container endpoint Delete a multi-container endpoint Use a multi-container endpoint with direct invocation Invoke a multi-container endpoint with direct invocation Security with multi-container endpoints with direct invocation Metrics for multi-container endpoints with direct invocation Autoscale multi-container endpoints Troubleshoot multi-container endpoints Ping Health Check Errors Missing accept-bind-to-port=true Docker label Deploy an Inference Pipeline Sample Notebooks for Inference Pipelines Feature Processing with Spark ML and Scikit-learn Feature Processing with Spark ML Feature Processing with Sci-kit Learn Create a Pipeline Model Run Real-time Predictions with an Inference Pipeline Create and Deploy an Inference Pipeline Endpoint Request Real-Time Inference from an Inference Pipeline Endpoint Realtime inference pipeline example Run Batch Transforms with Inference Pipelines Inference Pipeline Logs and Metrics Use Metrics to Monitor Multi-container Models Use Logs to Monitor an Inference Pipeline Troubleshoot Inference Pipelines Troubleshoot Amazon ECR Permissions for Inference Pipelines Use CloudWatch Logs to Troubleshoot SageMaker Inference Pipelines Use Error Messages to Troubleshoot Inference Pipelines Automatically Scale Amazon SageMaker Models Prerequisites Autoscaling policy overview Target metric for autoscaling Minimum and maximum capacity Cooldown period Permissions Service-linked role Configure model autoscaling with the console Register a model Register a model with the AWS CLI Register a model with the Application Auto Scaling API Define a scaling policy Use a predefined metric Use a custom metric Add a cooldown period Apply a scaling policy Apply a scaling policy (AWS CLI) Apply a scaling policy (Application Auto Scaling API) Edit a scaling policy Scale-in Disable scale-in activity Scale-out Disable scale-out activity Edit a scaling policy (Console) Edit a scaling policy (AWS CLI or Application Auto Scaling API) Delete a scaling policy Delete a scaling policy (Console) Delete a scaling policy (AWS CLI or Application Auto Scaling API) Delete a scaling policy (AWS CLI) Delete a scaling policy (Application Auto Scaling API) Query Endpoint Autoscaling History How To Query Endpoint Autoscaling Actions How to Identify Blocked AutoScaling Due to Instance Quotas Update or delete endpoints that use automatic scaling Update endpoints that use automatic scaling Delete endpoints configured for automatic scaling Load testing your autoscaling configuration Determine the performance characteristics Calculate the target load Use AWS CloudFormation to update autoscaling policies Host Instance Storage Volumes Test models in production Test models by specifying traffic distribution Test models by invoking specific variants Model A/B test example Step 1: Create and deploy models Step 2: Invoke the deployed models Step 3: Evaluate model performance Step 4: Increase traffic to the best model Troubleshoot Amazon SageMaker Model Deployments Detection Errors in the Active CPU Count Deployment Best Practices Deploy Multiple Instances Across Availability Zones Amazon SageMaker Model Monitor How Model Monitor Works Model Monitor Sample Notebooks Monitor Data Quality Create a Baseline Schema for Statistics (statistics.json file) CloudWatch Metrics Schema for Violations (constraint_violations.json file) Monitor Model Quality Create a Model Quality Baseline Schedule Model Quality Monitoring Jobs Ingest Ground Truth Labels and Merge Them With Predictions Model Quality Metrics Regression Metrics Binary Classification Metrics Multiclass Metrics Model Quality CloudWatch Metrics Monitor Bias Drift for Models in Production Model Monitor Sample Notebook Create a Bias Drift Baseline Schedule Bias Drift Monitoring Jobs Inspect Reports for Data Bias Drift Monitor Feature Attribution Drift for Models in Production Model Monitor Example Notebook Create a SHAP Baseline for Models in Production Schedule Feature Attribute Drift Monitoring Jobs Inspect Reports for Feature Attribute Drift in Production Models Capture Data Schedule Monitoring Jobs The cron Expression for Monitoring Schedule Amazon SageMaker Model Monitor Prebuilt Container Interpret Results List Executions Inspect a Specific Execution List Generated Reports Violations Report Visualize Results in Amazon SageMaker Studio Advanced Topics Customize Monitoring Preprocessing and Postprocessing Postprocessing Script Preprocessing Script Bring Your Own Containers Container Contract Inputs Container Contract Outputs Schema for Statistics (statistics.json file) Schema for Constraints (constraints.json file) CloudWatch Metrics for Bring Your Own Containers Create a Monitoring Schedule with an AWS CloudFormation Custom Resource Custom Resource Lambda Custom Resource Code Register and Deploy Models with Model Registry Create a Model Group Create a Model Package Group (Boto3) Create a Model Package Group (SageMaker Studio) Register a Model Version Register a Model Version (SageMaker Pipelines) Register a Model Version (Boto3) View Model Groups and Versions View a List of Model Versions in a Group View a List of Model Versions in a Group (Boto3) View a List of Model Versions in a Group (SageMaker Studio) View the Details of a Model Version View the Details of a Model Version (Boto3) View the Details of a Model Version (SageMaker Studio) Update the Approval Status of a Model Update the Approval Status of a Model (Boto3) Update the Approval Status of a Model (SageMaker Studio) Deploy a Model in the Registry Deploy a Model in the Registry (SageMaker SDK) Deploy a Model in the Registry (Boto3) Deploy a Model Version from a Different Account View the Deployment History of a Model Compile and Deploy Models with Neo What is SageMaker Neo? How it Works Neo Sample Notebooks Use Neo to Compile a Model Prepare Model for Compilation What input data shapes does SageMaker Neo expect? Keras MXNet/ONNX PyTorch TensorFlow TFLite XGBoost Saving Models for SageMaker Neo Keras MXNet PyTorch TensorFlow Built-In Estimators Compile a Model (AWS Command Line Interface) Compile a Model (Amazon SageMaker Console) Compile a Model (Amazon SageMaker SDK) Cloud Instances Supported Instance Types and Frameworks Cloud Instances Instance Types AWS Inferentia Amazon Elastic Inference Deploy a Model Prerequisites Deploy a Compiled Model Using SageMaker SDK If you compiled your model using the SageMaker SDK If you compiled your model using MXNet or PyTorch If you compiled your model using Boto3, SageMaker console, or the CLI for TensorFlow Deploy a Compiled Model Using Boto3 Deploy the Model Deploy a Compiled Model Using the AWS CLI Deploy the Model Create a Model Create an Endpoint Configuration Create an Endpoint Deploy a Compiled Model Using the Console Deploy the Model Request Inferences from a Deployed Service Request Inferences from a Deployed Service (Amazon SageMaker SDK) PyTorch and MXNet TensorFlow Request Inferences from a Deployed Service (Boto3) Request Inferences from a Deployed Service (AWS CLI) Inference Container Images Amazon SageMaker XGBoost TensorFlow MXNet PyTorch Edge Devices Supported Frameworks, Devices, Systems, and Architectures Supported Frameworks Supported Devices, Chip Architectures, and Systems Devices Systems and Chip Architectures Tested Models DarkNet MXNet Keras ONNX PyTorch (FP32) TensorFlow TensorFlow-Lite Deploy Models Deploy a Compiled Model (DLR) Deploy a Model (AWS IoT Greengrass) Getting Started with Neo on Edge Devices Prerequisites Step 1: Compile the Model Step 2: Set Up Your Device Step 3: Make Inferences on Your Device Troubleshoot Errors Error Classification Types Client permission error Load error Compilation error Troubleshoot Neo Compilation Errors How to Use This Page Framework-Related Errors TensorFlow Keras MXNet Infrastructure-Related Errors Troubleshoot Neo Inference Errors Troubleshoot Ambarella Errors Setting up the Configuration File Calibration Images Mean and Scale SageMaker Edge Manager Why Use Edge Manager? How Does it Work? How Do I Use SageMaker Edge Manager? Getting Started Setting Up Train, Compile, and Package Your Model Create and Register Fleets and Authenticate Devices Download and Set Up Edge Manager Run Agent Set Up Devices and Fleets Create a Fleet Create a Fleet (Boto3) Create a Fleet (Console) Register a Device Register a Device (Boto3) Register a Device (Console) Check Status Package Model Prerequisites Package a Model (Amazon SageMaker Console) Package a Model (Boto3) Edge Manager Agent Download and Set Up Edge Manager Agent Manually How the Agent Works Installing Edge Manager Agent Running SageMaker Edge Manager Agent Deploy Model Package and Edge Manager Agent with AWS IoT Greengrass Prerequisites Create AWS IoT Greengrass V2 Components Autogenerated Component Create a Hello World custom component Deploy Components to Your Device To deploy your components (console) To deploy your components (AWS CLI) Manage Model Load Model Unload Model List Models Describe Model Capture Data Get Capture Status Predict Using Docker containers with SageMaker Scenarios for Running Scripts, Training Algorithms, or Deploying Models with SageMaker Docker Container Basics Use Prebuilt SageMaker Docker images Prebuilt SageMaker Docker Images for Deep Learning Using the SageMaker Python SDK Extending Prebuilt SageMaker Docker Images Prebuilt Amazon SageMaker Docker Images for Scikit-learn and Spark ML Using the SageMaker Python SDK Specifying the Prebuilt Images Manually Finding Available Images Train a Deep Graph Network What Is a Deep Graph Network? Get Started Run a Graph Network Training Example Examples Use a Deep Learning Container with DGL Bring Your Own Container with DGL Extend a Prebuilt Container Requirements to Extend a Prebuilt Container Extend SageMaker Containers to Run a Python Script Step 1: Create an SageMaker Notebook Instance Step 2: Create and Upload the Dockerfile and Python Training Scripts Step 3: Build the Container Step 4: Test the Container Step 5: Push the Container to Amazon Elastic Container Registry (Amazon ECR) Step 6: Clean up Resources Adapting Your Own Docker Container to Work with SageMaker Individual Framework Libraries Using the SageMaker Training and Inference Toolkits SageMaker Toolkits Containers Structure Single Versus Multiple Containers Adapting Your Own Training Container Step 1: Create a SageMaker notebook instance Step 2: Create and upload the Dockerfile and Python training scripts Step 3: Build the container Step 4: Test the container Step 5: Push the container to Amazon Elastic Container Registry (Amazon ECR) Step 6: Clean up resources Adapting Your Own Inference Container Step 1: Create an Inference Handler The model_fn Function The input_fn Function The predict_fn Function The output_fn Function Step 2: Implement a Handler Service Step 3: Implement an Entrypoint Step 4: Write a Dockerfile Step 5: Build and Register Your Container Create a container with your own algorithms and models Use Your Own Training Algorithms How Amazon SageMaker Runs Your Training Image How Amazon SageMaker Provides Training Information Hyperparameters Environment Variables Input Data Configuration Training Data Distributed Training Configuration Run Training with EFA Prerequisites Install EFA and required packages Considerations when creating your container Verify that your EFA device is recognized Running a training job with EFA How Amazon SageMaker Signals Algorithm Success and Failure How Amazon SageMaker Processes Training Output Use Your Own Inference Code Use Your Own Inference Code with Hosting Services How SageMaker Runs Your Inference Image How SageMaker Loads Your Model Artifacts How Containers Serve Requests How Your Container Should Respond to Inference Requests How Your Container Should Respond to Health Check (Ping) Requests Use a Private Docker Registry for Real-Time Inference Containers Store Images in a Private Docker Registry other than Amazon Elastic Container Registry Use an Image from a Private Docker Registry for Real-time Inference Allow SageMaker to authenticate to a private Docker registry Create the Lambda function Give your execution role permission to Lambda Create an interface VPC endpoint for Lambda Use Your Own Inference Code with Batch Transform How SageMaker Runs Your Inference Image How SageMaker Loads Your Model Artifacts How Containers Serve Requests How Your Container Should Respond to Inference Requests How Your Container Should Respond to Health Check (Ping) Requests Example Notebooks: Use Your Own Algorithm or Model Setup Host Models Trained in Scikit-learn Package TensorFlow and Scitkit-learn Models for Use in SageMaker Train and Deploy a Neural Network on SageMaker Training Using Pipe Mode Bring Your Own R Model Extend a Prebuilt PyTorch Container Image Train and Debug Training Jobs on a Custom Container Troubleshooting your Docker containers SageMaker Workflows Amazon SageMaker Model Building Pipelines SageMaker Pipelines Overview Pipeline Structure Access Management Pipeline Role Permissions Pipeline Step Permissions Service Control Policies with Pipelines Pipeline Parameters Pipeline Steps Step Types Processing Step Training Step Tuning Step CreateModel Step RegisterModel Step Transform Step Condition Step Callback Step Lambda Step Step Properties Data Dependency Between Steps Custom Dependency Between Steps Use a Custom Image in a Step Property Files and JsonGet Caching Pipeline Steps Enabling Step Caching Amazon EventBridge Integration Schedule a Pipeline with Amazon EventBridge Prerequisites Create an EventBridge rule using the EventBridge console Create an EventBridge rule using the AWS CLI Amazon SageMaker Experiments Integration Default Behavior Disable Experiments Integration Specify a Custom Experiment Name Specify a Custom Trial Name SageMaker Pipelines Quotas Pipeline Quotas Executions Quotas Step Quotas Parameters Quotas Condition Step Quotas Property Files Quotas Metadata Quotas Troubleshooting Amazon SageMaker Model Building Pipelines Create and Manage SageMaker Pipelines Define a Pipeline Prerequisites Set Up Your Environment Create a Pipeline Step 1: Download the Dataset Step 2: Define Pipeline Parameters Step 3: Define a Processing Step for Feature Engineering Step 4: Define a Training step Step 5: Define a Processing Step for Model Evaluation Step 6: Define a CreateModelStep for Batch Transformation Step 7: Define a TransformStep to Perform Batch Transformation Step 8: Define a RegisterModel Step to Create a Model Package Step 9: Define a Condition Step to Verify Model Accuracy Step 10: Create a pipeline Run a pipeline Prerequisites Step 1: Start the Pipeline Step 2: Examine a Pipeline Execution Step 3: Override Default Parameters for a Pipeline Execution Step 4: Stop and Delete a Pipeline Execution View, Track, and Execute SageMaker Pipelines in SageMaker Studio View a Pipeline View a Pipeline Execution View Experiment Entities Created by SageMaker Pipelines Execute a Pipeline Track the Lineage of a SageMaker ML Pipeline Automate MLOps with SageMaker Projects What is a SageMaker Project? When Should You Use a SageMaker Project? Do I Need to Create a Project to Use SageMaker Pipelines? Why Should You Use MLOps? Challenges with MLOps Benefits of MLOps SageMaker Studio Permissions Required to Use Projects Create an MLOps Project using Amazon SageMaker Studio MLOps Project Templates Use SageMaker Provided Project Templates MLOps template for model building, training, and deployment MLOps template for model building, training, and deployment with third-party Git repositories using CodePipeline MLOps template for model building, training, and deployment with third-party Git repositories using Jenkins Update SageMaker Projects to Use Third-Party Git Repositories Create Custom Project Templates View Project Resources SageMaker MLOps Project Walkthrough Step 1: Create the Project Step 2: Clone the Code Repository Step 3: Make a Change in the Code Step 4: Approve the Model (Optional) Step 5: Deploy the Model Version to Production Step 6: Cleanup Resources Amazon SageMaker ML Lineage Tracking Tracking Entities Amazon SageMaker Created Tracking Entities Tracking Entities for SageMaker Jobs Tracking Entities for Model Packages Tracking Entities for Endpoints Manually Create Tracking Entities Manually Create Entities Manually Track a Workflow Limits Kubernetes Orchestration SageMaker Operators for Kubernetes What is an operator? Prerequisites Permissions overview IAM role-based setup and operator deployment Cluster-scoped deployment Create an OpenID Connect Provider for Your Cluster Get the OIDC ID Create an IAM Role Attach the AmazonSageMakerFullAccess Policy to the Role Deploy the Operator Deploy the Operator Using YAML Deploy the Operator Using Helm Charts Verify the operator deployment Namespace-scoped deployment Create an OpenID Connect Provider for Your Amazon EKS cluster Get your OIDC ID Create your IAM Role Attach the AmazonSageMakerFullAccess Policy to your Role Deploy the Operator to Your Namespace Deploy the Operator to Your Namespace Using YAML Deploy the Operator to Your Namespace Using Helm Charts Verify the operator deployment to your namespace Install the SageMaker logs kubectl plugin Delete operators Delete cluster-based operators Operators installed using YAML Operators installed using Helm Charts Delete namespace-based operators Operators installed with YAML Operators installed with Helm Charts Troubleshooting Debugging a Failed Job Deleting an Operator CRD Images and SMlogs in each Region Using Amazon SageMaker Jobs TrainingJob operator Create a TrainingJob Using a YAML File Create a TrainingJob Using a Helm Chart Create the TrainingJob Verify Your Training Helm Chart List TrainingJobs TrainingJob Status Values Secondary Status Values Describe a TrainingJob View Logs from TrainingJobs Delete TrainingJobs HyperParameterTuningJob operator Create a HyperparameterTuningJob Using a YAML File Create a HyperparameterTuningJob using a Helm Chart Create the HyperparameterTuningJob Verify Chart Installation List HyperparameterTuningJobs Hyperparameter Tuning Job Status Values Status Counters Best TrainingJob Spawned TrainingJobs Describe a HyperparameterTuningJob View Logs from HyperparameterTuningJobs Delete a HyperparameterTuningJob BatchTransformJob operator Create a BatchTransformJob Using a YAML File Create a BatchTransformJob Using a Helm Chart Get the Helm installer directory Configure the Helm Chart Create a BatchTransformJob List BatchTransformJobs Batch Transform Status Values Describe a BatchTransformJob View Logs from BatchTransformJobs Delete a BatchTransformJob HostingDeployment operator Configure a HostingDeployment Resource Create a HostingDeployment List HostingDeployments HostingDeployment Status Values Describe a HostingDeployment Invoking the Endpoint Update HostingDeployment Delete the HostingDeployment ProcessingJob operator Create a ProcessingJob Using a YAML File List ProcessingJobs Describe a ProcessingJob Delete a ProcessingJob HostingAutoscalingPolicy (HAP) Operator Create a HostingAutoscalingPolicy Using a YAML File Sample 1: Apply a Predefined Metric to a Single Endpoint Variant Sample 2: Apply a Custom Metric to a Single Endpoint Variant Sample 3: Apply a Scaling Policy to Multiple Endpoints and Variants Considerations for HostingAutoscalingPolicies for Multiple Endpoints and Variants List HostingAutoscalingPolicies Describe a HostingAutoscalingPolicy Update a HostingAutoscalingPolicy Delete a HostingAutoscalingPolicy Update or Delete an Endpoint with a HostingAutoscalingPolicy SageMaker Components for Kubeflow Pipelines What is Kubeflow Pipelines? Kubeflow Pipeline components What do SageMaker Components for Kubeflow Pipelines provide? Training components Inference components Ground Truth components IAM permissions Converting Pipelines to use SageMaker Using SageMaker Components Setup Set up a gateway node Set up an Amazon EKS cluster Install Kubeflow Pipelines Access the KFP UI Set up port forwarding to the KFP UI service Access the KFP UI service Create IAM Users/Roles for KFP pods and the SageMaker service Create a KFP execution role Create an SageMaker execution role Add access to additional IAM users or roles Running the Kubeflow Pipeline Prepare datasets Create a Kubeflow Pipeline using SageMaker Components Input Parameters Compile and deploy your pipeline Install KFP SDK Compile your pipeline Upload and run the pipeline using the KFP CLI Upload and run the pipeline using the KFP UI Running predictions Configure permissions to run predictions Run predictions View results and logs Cleanup Using Amazon Augmented AI for Human Review Get Started with Amazon Augmented AI Core Components of Amazon A2I Task Types Human Review Workflow (Flow Definition) Human Loops Prerequisites to Using Augmented AI Tutorial: Get Started in the Amazon A2I Console Prerequisites Step 1: Create a Work Team Step 2: Create a Human Review Workflow Step 3: Start a Human Loop Step 4: View Human Loop Status in Console Step 5: Download Output Data Tutorial: Get Started Using the Amazon A2I API Create a Private Work Team Create a Human Review Workflow Create a Human Task UI Create JSON to specify activation conditions Create a human review workflow Create a Human Loop Use Cases and Examples Using Amazon A2I Use SageMaker Notebook Instance with Amazon A2I Jupyter Notebook Use Amazon Augmented AI with Amazon Textract Get Started: Integrate a Human Review into an Amazon Textract Analyze Document Job End-to-End Example Using Amazon Textract and Amazon A2I A2I Textract Worker Console Preview Use Amazon Augmented AI with Amazon Rekognition Get Started: Integrate a Human Review into an Amazon Rekognition Image Moderation Job End-to-end Demo Using Amazon Rekognition and Amazon A2I A2I Rekognition Worker Console Preview Use Amazon Augmented AI with Custom Task Types End-to-end Tutorial Using Amazon A2I Custom Task Types Create a Human Review Workflow Create a Human Review Workflow (Console) Next Steps Create a Human Review Workflow (API) Next Steps JSON Schema for Human Loop Activation Conditions in Amazon Augmented AI Use Human Loop Activation Conditions JSON Schema with Amazon Textract ImportantFormKeyConfidenceCheck Inputs and Results MissingImportantFormKey Inputs and Results Sampling Inputs and Results Examples Use Human Loop Activation Conditions JSON Schema with Amazon Rekognition ModerationLabelConfidenceCheck Inputs Sampling Inputs Examples Delete a Human Review Workflow Delete a Flow Definition Using the Console or the SageMaker API Create and Start a Human Loop Create and Start a Human Loop for a Built-in Task Type Create an Amazon Textract Human Loop Create an Amazon Rekognition Human Loop Create and Start a Human Loop for a Custom Task Type Next Steps: Delete a Human Loop Human Loop Data Retention and Deletion Stop and Delete a Flow Definition Using the Console or the Amazon A2I API Create and Manage Worker Task Templates Create and Delete Worker Task Templates Create a Worker Task Template Delete a Worker Task Template Create Custom Worker Task Templates Develop Templates Locally Use External Assets Track Your Variables Custom Template Example for Amazon Textract Custom Template Example for Amazon Rekognition Add Automation with Liquid Use Variable Filters Autoescape and Explicit Escape escape_once skip_autoescape to_json grant_read_access Preview a Worker Task Template Creating Good Worker Instructions Create Good Worker Instructions Add Example Images to Your Instructions Monitor and Manage Your Human Loop Amazon A2I Output Data Output Data From Built-In Task Types Output Data From Custom Task Types Track Worker Activity Permissions and Security in Amazon Augmented AI CORS Permission Requirement Add Permissions to the IAM Role Used to Create a Flow Definition Create an IAM User That Can Invoke Amazon A2I API Operations Create an IAM User With Permissions to Invoke Amazon A2I, Amazon Textract, and Amazon Rekognition API Operations Enable Worker Task Template Previews Using Amazon A2I with AWS KMS Encrypted Buckets Additional Permissions and Security Resources Use Amazon CloudWatch Events in Amazon Augmented AI Send Events from Your Human Loop to CloudWatch Events Set Up a Target to Process Events Use Human Review Output More Information Use APIs in Amazon Augmented AI Programmatic Tutorials Buy and Sell Amazon SageMaker Algorithms and Models in AWS Marketplace Topics SageMaker Algorithms SageMaker Model Packages Use your own algorithms and models with the AWS Marketplace Create Algorithm and Model Package Resources Create an Algorithm Resource Create an Algorithm Resource (Console) Create an Algorithm Resource (API) Create a Model Package Resource Create a Model Package Resource (Console) Create a Model Package Resource (API) Use Algorithm and Model Package Resources Use an Algorithm to Run a Training Job Use an Algorithm to Run a Training Job (Console) Use an Algorithm to Run a Training Job (API) Use an Algorithm to Run a Training Job (Amazon SageMaker Python SDK) Use an Algorithm to Run a Hyperparameter Tuning Job Use an Algorithm to Run a Hyperparameter Tuning Job (Console) Use an Algorithm to Run a Hyperparameter Tuning Job (API) Use an Algorithm to Run a Hyperparameter Tuning Job (Amazon SageMaker Python SDK) Use a Model Package to Create a Model Use a Model Package to Create a Model (Console) Use a Model Package to Create a Model (API) Use a Model Package to Create a Model (Amazon SageMaker Python SDK) Sell Amazon SageMaker Algorithms and Model Packages Topics Develop Algorithms and Models in Amazon SageMaker Develop Algorithms in SageMaker Develop Models in SageMaker List Your Algorithm or Model Package on AWS Marketplace Find and Subscribe to Algorithms and Model Packages on AWS Marketplace Use Algorithms and Model Packages Security in Amazon SageMaker Access Control Access control and SageMaker Studio notebooks Control root access to a SageMaker notebook instance Data Protection in Amazon SageMaker Protect Data at Rest Using Encryption Studio notebooks Notebook instances and SageMaker jobs Protecting Data in Transit with Encryption Protect Communications Between ML Compute Instances in a Distributed Training Job Enable Inter-Container Traffic Encryption (API) Enable Inter-Container Traffic Encryption (Console) Enable Inter-container Traffic Encryption in a Training Job Enable Inter-container Traffic Encryption in a Hyperparameter Tuning Job Key Management Internetwork Traffic Privacy Identity and Access Management for Amazon SageMaker Audience Authenticating with Identities AWS account root user IAM Users and Groups IAM Roles Managing Access Using Policies Identity-Based Policies Resource-Based Policies Access Control Lists (ACLs) Other Policy Types Multiple Policy Types How Amazon SageMaker Works with IAM SageMaker Identity-Based Policies Actions Resources Condition Keys Examples SageMaker Resource-Based Policies Authorization Based on SageMaker Tags SageMaker IAM Roles Using Temporary Credentials with SageMaker Service-Linked Roles Service Roles Choosing an IAM Role in SageMaker Amazon SageMaker Identity-Based Policy Examples Policy Best Practices Using the SageMaker Console Permissions Required to Use the Amazon SageMaker Console Permissions Required to Use the Amazon SageMaker Ground Truth Console Permissions Required to Use the Amazon Augmented AI (Preview) Console Allow Users to View Their Own Permissions Control Creation of SageMaker Resources with Condition Keys Control Access to SageMaker Resources by Using File System Condition Keys Restrict an IAM User to Specific Directories and Access Modes Restrict an IAM User to a Specific File System Restrict Training to a Specific VPC Restrict Access to Workforce Types for Ground Truth Labeling Jobs and Amazon A2I Human Review Workflows Enforce Encryption of Input Data Enforce Encryption of Notebook Instance Storage Volume Enforce Network Isolation for Training Jobs Enforce a Specific Instance Type for Training Jobs Enforce a Specific EI Accelerator for Training Jobs Enforce Disabling Internet Access and Root Access for Creating Notebook Instances Control Access to the SageMaker API by Using Identity-based Policies Restrict Access to SageMaker API and Runtime to Calls from Within Your VPC Limit Access to SageMaker API and Runtime Calls by IP Address Limit Access to a Notebook Instance by IP Address Control Access to SageMaker Resources by Using Tags Require the Presence or Absence of Tags for API Calls SageMaker Roles Get execution role Add Additional Amazon S3 Permissions to an SageMaker Execution Role Passing Roles CreateAutoMLJob API: Execution Role Permissions CreateDomain API: Execution Role Permissions CreateImage and UpdateImage APIs: Execution Role Permissions CreateNotebookInstance API: Execution Role Permissions CreateHyperParameterTuningJob API: Execution Role Permissions CreateProcessingJob API: Execution Role Permissions CreateTrainingJob API: Execution Role Permissions CreateModel API: Execution Role Permissions AWS Managed Policies for Amazon SageMaker AmazonSageMakerFullAccess AmazonSageMakerReadOnly SageMaker Updates to AWS Managed Policies Amazon SageMaker API Permissions: Actions, Permissions, and Resources Reference Troubleshooting Amazon SageMaker Identity and Access I Am Not Authorized to Perform an Action in SageMaker I Am Not Authorized to Perform iam:PassRole I Want to View My Access Keys I'm an Administrator and Want to Allow Others to Access SageMaker I Want to Allow People Outside of My AWS Account to Access My SageMaker Resources Logging and Monitoring Compliance Validation for Amazon SageMaker Resilience in Amazon SageMaker Infrastructure Security in Amazon SageMaker SageMaker Scans AWS Marketplace Training and Inference Containers for Security Vulnerabilities Connect to Resources in a VPC Connect SageMaker Studio Notebooks to Resources in a VPC Default communication with the internet VPC only communication with the internet Requirements to use VPC only mode Connect a Notebook Instance to Resources in a VPC Default communication with the internet VPC communication with the internet Security and Shared Notebook Instances Run Training and Inference Containers in Internet-Free Mode Network Isolation Network isolation with a VPC Connect to SageMaker Through a VPC Interface Endpoint Create a VPC Endpoint Policy for SageMaker Create a VPC Endpoint Policy for Amazon SageMaker Feature Store Connect to SageMaker Studio Through an Interface VPC Endpoint Create a VPC Endpoint Policy for SageMaker Studio Allow Access Only from Within Your VPC Connect to a Notebook Instance Through a VPC Interface Endpoint Connect Your Private Network to Your VPC Create a VPC Endpoint Policy for SageMaker Notebook Instances Restrict Access to Connections from Within Your VPC Connect Your Private Network to Your VPC Give SageMaker Access to Resources in your Amazon VPC Give SageMaker Processing Jobs Access to Resources in Your Amazon VPC Configure a Processing Job for Amazon VPC Access Configure Your Private VPC for SageMaker Processing Ensure That Subnets Have Enough IP Addresses Create an Amazon S3 VPC Endpoint Use a Custom Endpoint Policy to Restrict Access to S3 Restrict Package Installation on the Processing Container Configure Route Tables Configure the VPC Security Group Connect to Resources Outside Your VPC Give SageMaker Training Jobs Access to Resources in Your Amazon VPC Configure a Training Job for Amazon VPC Access Configure Your Private VPC for SageMaker Training Ensure That Subnets Have Enough IP Addresses Create an Amazon S3 VPC Endpoint Use a Custom Endpoint Policy to Restrict Access to S3 Restrict Package Installation on the Training Container Configure Route Tables Configure the VPC Security Group Connect to Resources Outside Your VPC Give SageMaker Hosted Endpoints Access to Resources in Your Amazon VPC Configure a Model for Amazon VPC Access Configure Your Private VPC for SageMaker Hosting Ensure That Subnets Have Enough IP Addresses Create an Amazon S3 VPC Endpoint Use a Custom Endpoint Policy to Restrict Access to Amazon S3 Restrict Package Installation on the Model Container with a Custom Endpoint Policy Add Permissions for Endpoint Access for Containers Running in a VPC to Custom IAM Policies Configure Route Tables Connect to Resources Outside Your VPC Give Batch Transform Jobs Access to Resources in Your Amazon VPC Configure a Batch Transform Job for Amazon VPC Access Configure Your Private VPC for SageMaker Batch Transform Ensure That Subnets Have Enough IP Addresses Create an Amazon S3 VPC Endpoint Use a Custom Endpoint Policy to Restrict Access to S3 Restrict Package Installation on the Model Container Configure Route Tables Configure the VPC Security Group Connect to Resources Outside Your VPC Give Amazon SageMaker Clarify Jobs Access to Resources in Your Amazon VPC Configure a SageMaker Clarify Job for Amazon VPC Access SageMaker Clarify Job Amazon VPC Subnets and Security Groups Configure a Model Amazon VPC for Inference Configure Your Private Amazon VPC for SageMaker Clarify jobs Connect to Resources Outside Your Amazon VPC Configure the Amazon VPC Security Group Give SageMaker Compilation Jobs Access to Resources in Your Amazon VPC Configure a Compilation Job for Amazon VPC Access Configure Your Private VPC for SageMaker Compilation Ensure That Subnets Have Enough IP Addresses Create an Amazon S3 VPC Endpoint Use a Custom Endpoint Policy to Restrict Access to S3 Add Permissions for Compilation Job Running in a Amazon VPC to Custom IAM Policies Configure Route Tables Configure the VPC Security Group Monitor Amazon SageMaker Monitor Amazon SageMaker with Amazon CloudWatch SageMaker Endpoint Invocation Metrics SageMaker Multi-Model Endpoint Metrics SageMaker Jobs and Endpoint Metrics SageMaker Ground Truth Metrics SageMaker Feature Store Metrics SageMaker Pipelines Metrics Log Amazon SageMaker Events with Amazon CloudWatch Log Amazon SageMaker API Calls with AWS CloudTrail SageMaker Information in CloudTrail Operations Performed by Automatic Model Tuning Understanding SageMaker Log File Entries Automating Amazon SageMaker with Amazon EventBridge Training job state change Hyperparameter tuning job state change Transform job state change Endpoint state change Feature group state change Model package state change Pipeline execution state change Pipeline step state change SageMaker image state change SageMaker image version state change Availability Zones API Reference Guide for Amazon SageMaker Overview Programming Model for Amazon SageMaker Document History for Amazon SageMaker AWS glossary