Amazon SageMaker Python SDK

Amazon SageMaker Python SDK is an open source library for training and deploying machine-learned models on Amazon SageMaker.

With the SDK, you can train and deploy models using popular deep learning frameworks, algorithms provided by Amazon, or your own algorithms built into SageMaker-compatible Docker images.

Here you’ll find an overview and API documentation for SageMaker Python SDK. The project homepage is in Github: https://github.com/aws/sagemaker-python-sdk, where you can find the SDK source and installation instructions for the library.

Overview

• Using the SageMaker Python SDK
  • Train a Model with the SageMaker Python SDK
  • Using Models Trained Outside of Amazon SageMaker
  • Use Built-in Algorithms with Pre-trained Models in SageMaker Python SDK
  • SageMaker Automatic Model Tuning
  • SageMaker Asynchronous Inference
  • SageMaker Serverless Inference
  • SageMaker Batch Transform
  • Local Mode
  • Secure Training and Inference with VPC
  • Secure Training with Network Isolation (Internet-Free) Mode
  • Inference Pipelines
  • SageMaker Workflow
  • SageMaker Model Building Pipeline
  • SageMaker Model Monitoring
  • SageMaker Debugger
  • SageMaker Processing
  • Configuring and using defaults with the SageMaker Python SDK
  • Run Machine Learning code on SageMaker using remote function
  • FAQ
• Use Version 2.x of the SageMaker Python SDK
  • Installation
  • Breaking Changes
  • Non-Breaking Changes
  • Automatically Upgrade Your Code

The SageMaker Python SDK APIs:

• APIs
  • Feature Store APIs
  • Training APIs
  • Distributed Training APIs
  • Inference APIs
  • Governance APIs
  • Utility APIs

Frameworks

The SageMaker Python SDK supports managed training and inference for a variety of machine learning frameworks:

• Frameworks
  • Apache MXNet
  • Chainer
  • Hugging Face
  • PyTorch
  • Reinforcement Learning
  • Scikit-Learn
  • SparkML Serving
  • TensorFlow
  • XGBoost
  • Deep Java Library (DJL)

SageMaker Built-in Algorithms

Amazon SageMaker provides implementations of some common machine learning algorithms optimized for GPU architecture and massive datasets.

• Built-in Algorithms
  • Amazon Estimators
  • Tabular
  • Text
  • Time-series
  • Unsupervised
  • Vision

Workflows

Orchestrate your SageMaker training and inference workflows with Airflow and Kubernetes.

• Workflows
  • SageMaker Pipelines
  • Airflow Workflows
  • AWS Step Functions
  • SageMaker Lineage

Amazon SageMaker Experiments

You can use Amazon SageMaker Experiments to track machine learning experiments.

• Amazon SageMaker Experiments
  • Experiments

Amazon SageMaker Debugger

You can use Amazon SageMaker Debugger to automatically detect anomalies while training your machine learning models.

• Amazon SageMaker Debugger
  • Background
  • Setup
  • Capture real-time debugging data during model training in Amazon SageMaker
  • Continuous analyses through rules
  • Capture real-time TensorBoard data from the debugging hook
  • Interactive analysis using SageMaker Debugger SDK and visualizations
  • Default behavior and opting out
  • Learn More

Amazon SageMaker Feature Store

You can use Feature Store to store features and associated metadata, so features can be discovered and reused.

• Amazon SageMaker Feature Store

Amazon SageMaker Model Monitoring

You can use Amazon SageMaker Model Monitoring to automatically detect concept drift by monitoring your machine learning models.

• Amazon SageMaker Model Monitor
  • Background
  • Setup
  • Capture real-time inference data from Amazon SageMaker endpoints
  • Baselining and continuous monitoring
  • Visualize results
  • Delete the resources
  • Learn More

Amazon SageMaker Processing

You can use Amazon SageMaker Processing to perform data processing tasks such as data pre- and post-processing, feature engineering, data validation, and model evaluation

• Amazon SageMaker Processing
  • Background
  • Setup
  • Data Pre-Processing and Model Evaluation with scikit-learn
  • Data Processing with Spark
  • Learn More

Amazon SageMaker Model Building Pipeline

You can use Amazon SageMaker Model Building Pipelines to orchestrate your machine learning workflow.

• Amazon SageMaker Model Building Pipeline
  • Background
  • Code Examples
  • Example Notebooks
  • Limitations
  • Pipeline Composition Alternative