Skip to content

Releases: aws/sagemaker-python-sdk

v2.75.1

08 Feb 05:16
Compare
Choose a tag to compare

Bug Fixes and Other Changes

  • Add CMH to the non-P3 list

v2.75.0

05 Feb 05:59
Compare
Choose a tag to compare

Features

  • JumpStart Integration
  • Adds support for async inference
  • Update instance types for integ test

Bug Fixes and Other Changes

  • Revert "feature: CompilationStep support for Sagemaker Pipelines
  • gpu use p3/p2 per avail for region
  • jumpstart typo
  • pin pytest-xdist to avoid release failures
  • set sagemaker_connection and image_uri in register method
  • update to incorporate black v22, pin tox versions
  • Add deprecation warning in Clarify DataConfig

Documentation Changes

  • Jumpstart doc strings and added new sections
  • Add Jumpstart support documentation

v2.74.0

26 Jan 17:55
Compare
Choose a tag to compare

Features

  • Add support for SageMaker lineage queries context

Bug Fixes and Other Changes

  • support specifying a facet by its column index

Documentation Changes

  • more documentation for serverless inference

v2.73.0

19 Jan 02:16
Compare
Choose a tag to compare

Features

  • Add EMRStep support in Sagemaker pipeline
  • Adds Lineage queries in artifact, context and trial components
  • Add support for SageMaker lineage queries in action
  • Adds support for Serverless inference
  • support checkpoint to be passed from estimator
  • support JsonGet/Join parameterization in tuning step Hyperparameters
  • Support model pipelines in CreateModelStep
  • enable python 3.9
  • Add models_v2 under lineage context

Bug Fixes and Other Changes

  • allow kms_key to be passed for processing step
  • Remove duplicate vertex/edge in query lineage
  • update pricing link
  • Update CHANGELOG.md
  • fixes unnecessary session call while generating pipeline definition for lambda step

Documentation Changes

  • Enhance smddp 1.2.2 doc
  • Document the available ExecutionVariables

v2.72.3

10 Jan 19:39
Compare
Choose a tag to compare

Bug Fixes and Other Changes

  • update master from dev

v2.72.2

06 Jan 23:53
Compare
Choose a tag to compare

Bug Fixes and Other Changes

  • Update CHANGELOG.md
  • Increment static lineage pipeline
  • fix kmeans test deletion sequence, increment lineage statics
  • Re-deploy static integ test endpoint if it is not found
  • Add ContentType in test_auto_ml_describe
  • Model Registration with BYO scripts

Documentation Changes

  • SageMaker model parallel library 1.6.0 API doc

v2.72.1

20 Dec 16:35
Compare
Choose a tag to compare

Bug Fixes and Other Changes

  • typos and broken link
  • S3Input - add support for instance attributes
  • Prevent repack_model script from referencing nonexistent directories
  • Set ProcessingStep upload locations deterministically to avoid c…

v2.72.0

13 Dec 20:51
Compare
Choose a tag to compare

Features

  • allow conditional parellel builds

Bug Fixes and Other Changes

  • local mode - support relative file structure
  • fix endpoint bug

v2.71.0

06 Dec 16:33
Compare
Choose a tag to compare

Features

  • Add support for TF 2.6
  • Adding PT 17/18 Repo
  • Add profile_name support for Feature Store ingestion

Bug Fixes and Other Changes

  • Fix non-existent variable name
  • Add TF 2.6.2 on training
  • Recreate static lineage test data

v2.70.0

02 Dec 05:36
Compare
Choose a tag to compare

Features

  • update boto3 minor version >= 1.20.18
  • Add support for SageMaker lineage queries
  • add CV shap explainability for SageMaker Clarify
  • add NLP support for SageMaker Clarify
  • Add support for ModelMonitor/Clarify integration in model building pipelines
  • adding support for transformers 4.11 for SM Training Compiler
  • SM Training Compiler with an UI to enable/disable compilation for HuggingFace DLCs to speedup training

Bug Fixes and Other Changes

  • pin coveragepy
  • Add support for PyTorch 1.9.1
  • Update s3 path of scheduling analysis config on ClarifyCheckStep
  • documentation/logging to indicate correct place for DEBUG artifacts from SM trcomp
  • validate requested transformers version and use the best available version
  • Install custom pkgs