TensorFlow 2.1.0 will include breaking changes: First release candidate available

TensorFlow 2.1.0-rc0 is now available, the first release candidate for the machine learning platform’s next minor version. As Python 2.7 will reach end of life on January 1, 2020, TensorFlow 2.1 will be the last version to support it.

TensorFlow is an open source software library for ML that was originally developed by the Google Brain team in 2015. It has since become very popular within the open source community and was found to be the 5th most popular open source project on GitHub in the latest State of the Octoverse report.

SEE ALSO: Machine learning: TensorFlow Training on the JVM

In the release candidate for TensorFlow 2.1.0, new features have been added for the tensorflow pip package, Keras and more. Among the breaking changes are API renamings as well as removals, and six APIs are now stable.

Breaking Changes

The release candidate for 2.1.0 removes id from tf.Tensor.__repr__() as well as Operation.traceback_with_star. These six APIs have graduated from the experimental stage:

• tf.config.list_logical_devices
• tf.config.list_physical_devices
• tf.config.get_visible_devices
• tf.config.set_visible_devices
• tf.config.get_logical_device_configuration
• tf.config.set_logical_device_configuration

tf.config.experimentalVirtualDeviceConfiguration has been renamed tf.config.LogicalDeviceConfiguration

and tf.config.experimental_list_devices has been removed in favor of tf.config.list_logical_devices.

Features

The tensorflow pip package has received an update: GPU support is now included by default for Linux and Windows on machines with and without NVIDIA GPUs.

In TensorFlow 2.1.0-rc0, the deep learning platform NVIDIA TensorRT 6 is supported and enabled by default. According to NVIDIA, TensorRT 6 accelerates use cases like conversational AI and can be used within different frameworks like TensorFlow and ONNX Runtime.

SEE ALSO: Adopt these 10 strategic technologies in 2020

Further new features were added for tf.keras and tf.data. Until a known issue with building on Windows has been resolved, eigen strong lining has been turned off for Windows builds, which is expected to cause a slight reduction in speed.

See the release notes for more information on the changes in TensorFlow 2.1.0-rc0.