NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
PyPI, #999 most downloaded on PyPI
Open source library for training and deploying models on Amazon SageMaker.
Latest stable
3.29.0
Last release 17 days ago
05 Nov 2026
Ships on a steady schedule
a new release about every 2 weeks
Nearly every release is documented
notes for 60 of the last 60 stable releases
3 versions withdrawn
withdrawn after publishing
9 years old
699 releases · first in 2017
Drag across the ruler to compare, or tap a release to open it.
Enable customizing artifact output path
Add support for tags in to_pipeline API for feature processor
Your coding agent can read these notes before it upgrades. Set up the MCP server →
Support uncompressed model upload
add SageMaker FeatureStore feature processing
Remove deprecated option.s3url in favor of option.model_id
inf2 support to HuggingFaceModelAdd support for Deployment Recommendation ID in model.deploy(). No tagging support
SDK Defaults - Environment Variables
Add huggingface-llm 0.8.2 dlc images
Add tagging support for create ir job
Add huggingface-llm 0.6.0 dlc images
remove unnecessary log messages for loading existing experiment runs
Add TF Serving 2.12.1 images to the SM PySDK
Enable default role for Spark processors
Handle use case where endpoint is created outside of python …
Partition support for DJLModel using SM Training job
Add support for SageMaker Serverless inference Provisioned Concurrency feature
Add integ tests for remote_function, auto_capture functionality
Remove deprecated update_endpoint from deploy() args in TensorFlowModel
add support for lineage visualization using pyvis
Update Transformers 4.26 - TensorFlow 2.11.0 Image URI
Introduce TensorBoard app class
update the doc for Join function
[huggingface] Add torch.distributed support for Trainium and torchrun
torch.distributed support for Trainium and torchrunsupport different types of deletion mode
Added ml.p4d and ml.p4de as supported instances for DeepSpeed
Add support for JSON model inputs for Clarify Processor
add support for async inline error notifications
Deprecation warning for framework profiling for TF 2.12 and on, PT 2.0 and on
Support for SageMaker SDK Defaults
combined model + script artifact
AutoGluon 0.7.0 image_uris update
Fix cross account register model
[Feature - Hugging Face] Update Transformers 4.26 - PyTorch 1.13.1 Image uri
Add XGBoost framework 1.7-1 version
Add new region support for MX, PT, TF on SM Training
support JSON for input dataset and model output
with_feature_group [feature_store]
update feature store dataset builder docs
Revert back to stable apache-airflow-providers-amazon from 7.2.1 to 4.0.0.
Add DLC accounts for MEL Region
local mode deletion of temp files on job end
Add python 3.9 and spark 3.2 support for spark processor
feature store with_feature_group functionality changes
support cluster lifecycle management for Sagemaker EMR step
Add a deprecation note for DetailedProfileConfig
Display file diff on black-check
Adding image_uri config for DJL containers
add p2 deprecation for PT>=1.13
right_size() for inference recommender
tensorflow inference 2.10.1 release
Do not specify S3 path for disabled profiler
add RandomSeed to support reproducible HPO
Add p4de to smddp supported instance types
Add support for TF2.9.2 training images
Nothing published for this version
Feature Store dataset builder, delete_record, get_record, list_feature_group
Update for Tensorflow Serving 2.11 inference DLCs
Pop out ModelPackageName from pipeline definition
Algorithms Region Expansion OSU/DXB
Add Neo image uri config for Pytorch 1.12
One column per quarter.