Releases: aws/sagemaker-python-sdk
Releases · aws/sagemaker-python-sdk
v1.42.1
Bug fixes and other changes
- lazy import of tensorflow module
- skip airflow_config tests as they're blocking the release build
- skip lda tests in regions that does not support it.
- add airflow_config tests to canaries
- use correct STS endpoint for us-iso-east-1
v1.42.0
Features
- add estimator preparation to airflow configuration
Bug fixes and other changes
- correct airflow workflow for BYO estimators.
v1.41.0
Features
- enable sklearn for network isolation mode
v1.40.2
Bug fixes and other changes
- use new ECR images in us-iso-east-1 for TF and MXNet
v1.40.1
Bug fixes and other changes
- expose kms_key parameter for deploying from training and hyperparameter tuning jobs
Documentation changes
- Update sklearn default predict_fn
v1.40.0
Features
- add support to TF 1.14 serving with elastic accelerator.
v1.39.4
Bug fixes and other changes
- pass enable_network_isolation when creating TF and SKLearn models
v1.39.3
Bug fixes and other changes
- expose vpc_config_override in transformer() methods
- use Estimator.create_model in Estimator.transformer
v1.39.2
Bug fixes and other changes
- pass enable_network_isolation in Estimator.create_model
- use p2 instead of p3 for the Horovod test
v1.39.1
Bug fixes and other changes
- copy dependencies into new folder when repacking model
- make get_caller_identity_arn get role from DescribeNotebookInstance
- add https to regional STS endpoint
- clean up git support integ tests