v2.1.0

Features

• add Dockerfiles for PyTorch 1.5.0

v2.0.0

Breaking Changes

• Replace sagemaker-containers with sagemaker-training

v1.3.3

Bug Fixes and Other Changes

• change miniconda installation in 1.4.0 Dockerfiles

Testing and Release Infrastructure

• parallelize SageMaker integ test runs
• remove (unused) model_fn from training scripts

v1.3.2

Bug Fixes and Other Changes

• bump smdebug version

Testing and Release Infrastructure

• add requirements.txt integ test

v1.3.1

Bug Fixes and Other Changes

• upgrade pillow etc. to fix safety issues
• Upgrade sagemaker-containers and test with more than 1 epoch

v1.3.0

Features

• Install toolkit from PyPI.

Bug Fixes and Other Changes

• upgrade sagemaker-containers to 2.8.2
• Install jupyter_client 5.3.4 in advanced for py2 gpu image
• update smdebug

Testing and Release Infrastructure

• run test-toolkit unit tests for release
• run build steps only when necessary.
• refactor toolkit tests.

v1.2.4

Bug Fixes and Other Changes

• install sm experiments always when python 3.6 or greater

v1.2.3

Bug Fixes and Other Changes

• Update smdebug to 0.7.0
• install sagemaker-experiments package only for 3.6

v1.2.2

Bug Fixes and Other Changes

• upgrade to latest sagemaker-experiments
• install SageMaker Python SDK into Python 3 images

v1.2.1

Bug Fixes and Other Changes

• Install awscli from pypi instead of conda for PyTorch containers