Skip to content

Releases: aws/sagemaker-python-sdk

v2.29.1

09 Mar 18:20

Choose a tag to compare

Bug Fixes and Other Changes

  • return all failed row indices in feature_group.ingest
  • move service-role path parsing for AmazonSageMaker-ExecutionRole for get_execution_role() into except block of IAM get_role() call and add warning message
  • add description parameter for RegisterModelStep
  • add type annotations for Lineage

Documentation Changes

  • remove ellipsis from CHANGELOG.md

v2.29.0

04 Mar 18:51

Choose a tag to compare

Features

  • add support for TensorFlow 2.4.1 for training, inference and…
  • Support profiler config in the pipeline training job step
  • support PyTorch 1.7.1 training, inference and data parallel

v2.28.0

03 Mar 16:28

Choose a tag to compare

Features

  • support creating endpoints with model images from private registries

v2.27.1

03 Mar 03:03

Choose a tag to compare

Bug Fixes and Other Changes

  • Change Estimator.logs() to use latest_training_job.name
  • mask creds from docker commands in local mode. Closes #2118

Documentation Changes

  • fix pipelines processing step typo
  • remove double 'enable-network-isolation' description

v2.27.0

01 Mar 21:33

Choose a tag to compare

Features

  • add inference_id to predict

Bug Fixes and Other Changes

  • disable profiler by default for regions not support it

Documentation Changes

  • add TF 2.4.1 support to sm distributed data parallel docs and other updates

v2.26.0

26 Feb 19:36

Choose a tag to compare

Features

  • Add Framework Version support for PyTorch compilation (Neo)

Bug Fixes and Other Changes

  • add mxnet 1.7.0 eia configuration
  • update source constructor for lineage action and artifact

Documentation Changes

  • fix typo in create_monitoring_schedule method

v2.25.2

25 Feb 03:53

Choose a tag to compare

Bug Fixes and Other Changes

  • Use the output path to store the Clarify config file
  • feature group should ignore nan values
  • ignore failing smdataparallel test
  • Add tests for Training job & Transform job in visualizer
  • visualizer for pipeline processing job steps

Documentation Changes

  • update doc for Elastic Inference MXNet 1.7.0

v2.25.1

20 Feb 00:06

Choose a tag to compare

Bug Fixes and Other Changes

  • Add tests for visualizer to improve test coverage

Documentation Changes

  • specify correct return type

Testing and Release Infrastructure

  • rename canary_quick pytest mark to release

v2.25.0

19 Feb 06:48

Choose a tag to compare

Features

  • Enable step caching
  • Add other Neo supported regions for Inferentia inference images

Bug Fixes and Other Changes

  • remove FailStep from pipelines
  • use sagemaker_session in workflow tests
  • use ECR public for multidatamodel tests
  • add the mapping from py3 to cuda11 images
  • Add 30s cap time for tag tests
  • add build spec for slow tests
  • mark top 10 slow tests
  • remove slow test_run_xxx_monitor_baseline tests
  • pin astroid to 2.4.2

Testing and Release Infrastructure

  • unmark more flaky integ tests
  • remove canary_quick pytest mark from flaky/unnecessary tests
  • remove python3.8 from buildspec
  • remove py38 tox env
  • fix release buildspec typo
  • unblock regional release builds
  • lower test TPS for experiment analytics
  • move package preparation and publishing to the deploy step

v2.24.5

12 Feb 18:54

Choose a tag to compare

Bug Fixes and Other Changes

  • test_tag/test_tags method assert fix in association tests

Documentation Changes

  • removing mention of TF 2.4 from SM distributed model parallel docs
  • adding details about mpi options, other small updates