Releases: aws/sagemaker-python-sdk
Releases · aws/sagemaker-python-sdk
v1.51.4
Bug Fixes and Other Changes
- Check that session is a LocalSession when using local mode
- add tflite to Neo-supported frameworks
- ignore tags with 'aws:' prefix when creating an EndpointConfig based on an existing one
- allow custom image when calling deploy or create_model with various frameworks
Documentation Changes
- fix description of default model_dir for TF
- add more details about PyTorch eia
v1.51.3
Bug Fixes and Other Changes
- make repack_model only removes py file when new entry_point provided
v1.51.2
Bug Fixes and Other Changes
- handle empty inputs/outputs in ProcessingJob.from_processing_name()
- use DLC images for GovCloud
Testing and Release Infrastructure
- generate test job name at test start instead of module start
v1.51.1
Bug Fixes and Other Changes
- skip pytorch ei test in unsupported regions
Documentation Changes
- correct MultiString/MULTI_STRING docstring
v1.51.0
Features
- pytorch 1.3.1 eia support
Documentation Changes
- Update Kubernetes Operator default tag
- improve docstring for tuner.best_estimator()
v1.50.18.post0
Documentation Changes
- correct Estimator code_location default S3 path
v1.50.18
Bug Fixes and Other Changes
- change default compile model max run to 15 mins
v1.50.17.post0
Testing and Release Infrastructure
- fix PR builds to run on changes to their own buildspecs
- programmatically determine partition based on region
v1.50.17
Bug Fixes and Other Changes
- upgrade framework versions
v1.50.16
Bug Fixes and Other Changes
- use sagemaker_session when initializing Constraints and Statistics
- add sagemaker_session parameter to DataCaptureConfig
- make AutoML.deploy use self.sagemaker_session by default
Testing and Release Infrastructure
- unset region during integ tests
- use sagemaker_session fixture in all Airflow tests
- remove remaining TF legacy mode integ tests