Releases: aws/sagemaker-tensorflow-training-toolkit
Releases · aws/sagemaker-tensorflow-training-toolkit
v20.4.1
v20.4.0
Features
- Add heterogeneous cluster changes
v20.3.1
Bug Fixes and Other Changes
- Fix/ci
v20.3.0
Features
- use tensorflow 2.3.1 and add data parallel integ test
Bug Fixes and Other Changes
- upgrade to sagemaker-training 3.7.1
Testing and Release Infrastructure
- include granular buildspecs for dlc and generic cpu and gpu testing
v20.2.0
Features
- include sm-data-distributed and upgrade dependencies
v20.1.5
Bug Fixes and Other Changes
- workaround to print stderr when capture_error is True
v10.1.8
Bug Fixes and Other Changes
- workaround to print stderr when capture_error is True
v20.1.4
Bug Fixes and Other Changes
- propagate log level
v10.1.7
Bug Fixes and Other Changes
- propagate log level
v20.1.3
Bug Fixes and Other Changes
- add condition to avoid error when 'model_dir' is None