What's new and changed in Watson Machine Learning Accelerator

Watson™ Machine Learning Accelerator updates can include new features, bug fixes, and security updates. Updates are listed in reverse chronological order so that the latest release is at the beginning of the topic.

You can see a list of the new features for the platform and all of the services at What's new in IBM Cloud Pak® for Data.

Installing or upgrading Watson Machine Learning Accelerator

Ready to install or upgrade Watson Machine Learning Accelerator? See:

Related documentation:

Cloud Pak for Data Version 4.6.5

A new version of Watson Machine Learning Accelerator was released in 4.6.5 with Cloud Pak for Data 4.6.5.

Operand version: 3.5.0

Version 3.5.0 of the Watson Machine Learning Accelerator service includes various fixes.

Cloud Pak for Data Version 4.6.4

A new version of Watson Machine Learning Accelerator was released in 4.6.4 with Cloud Pak for Data 4.6.4.

Operand version: 3.4.0

Version 3.4.0 of the Watson Machine Learning Accelerator service includes various fixes.

Cloud Pak for Data Version 4.6.3

A new version of Watson Machine Learning Accelerator was released in February 2023 with Cloud Pak for Data 4.6.3.

Operand version: 3.3.0

This release includes the following changes:

New features

The 3.3.0 release of Watson Machine Learning Accelerator includes the following features and updates:

Train Deep Learning experiments with CPU resources: You can now specify CPU as an alternative to GPU as a resource type when defining a Deep Learning experiment. You can then assign the number of workers to apply for the experiment. For details, see Training neural networks using the deep learning experiment builder.

Security fixes

Version 3.3.0 of the Watson Machine Learning Accelerator service includes various fixes.

Cloud Pak for Data Version 4.6.1

A new version of Watson Machine Learning Accelerator was released in December 2022 with Cloud Pak for Data 4.6.1.

Operand version: 3.1.0

This release includes the following changes:

New features

The 3.1.0 release of Watson Machine Learning Accelerator includes the following features and updates:

Python 3.10.6: The Watson Machine Learning Accelerator service now uses Python 3.10.6 for running deep learning frameworks.
If you have existing models, be sure to update and test your models to use the latest supported frameworks for Python 3.10.6. For details, see Supported deep learning frameworks in the Watson Machine Learning Accelerator documentation.

Security fixes

Version 3.1.0 of the Watson Machine Learning Accelerator service includes various fixes.

Cloud Pak for Data Version 4.6.0

A new version of Watson Machine Learning Accelerator was released in November 2022 with Cloud Pak for Data 4.6.0.

Operand version: 3.0.0

This release includes the following changes:

New features

Version 3.0.0 of the Watson Machine Learning Accelerator service includes the following features and updates:

New deep learning libraries

You can now use the following deep learning libraries with Watson Machine Learning Accelerator:

TensorFlow 2.9.2
PyTorch 1.12.1
NVIDIA CUDA Toolkit 11.4.4
Python 3.10.4

New NVIDIA GPU Operator version

You can now use Watson Machine Learning Accelerator with the following versions of the NVIDIA GPU Operator:

x86-64

On OpenShift® 4.8, use NVIDIA GPU Operator 1.10 or 1.9
On OpenShift 4.10, use NVIDIA GPU Operator 1.10 or 1.9

For details, see: Installing the NVIDIA GPU Operator on OpenShift.

Power®

On OpenShift 4.8 and 4.10, use Rocket Software GPU Operator v22.9.0. For details, see: GPU Operator for POWER.

Customer-reported issues fixed in this release

DT135132: Users will logout if they press F5 button in Cloud Pak for Data (CPD) Watson Machine Learning Accelerator (WMLA) web console

Security fixes

Version 3.0.0 of the Watson Machine Learning Accelerator service includes various fixes.