Releases: aws-neuron/aws-neuron-sdk
Neuron SDK Release - February 16, 2022
Neuron 1.17.1 is a patch release. This release fixes a bug in TensorFlow Neuron that caused a memory leak. The memory leak was approximately 128b for each inference and exists in all versions of TensorFlow Neuron versions part of Neuron 1.16.0 to Neuron 1.17.0 releases. see Previous Releases Content for exact versions included in each release. This release only fixes the memory leak for TensorFlow versions 1.15 and 2.5 from Neuron. The other versions of TensorFlow Neuron will be fixed in a shortly upcoming release.
Neuron SDK Release - January 20, 2022
Neuron 1.17.0 release introduces the support of PyTorch 1.10, Tensorflow 2.5 update to version 2.5.2, new operators support in PyTorch and TensorFlow 1.15, in addition to enhancements and bug fixes in PyTorch, TensorFlow, MxNet, Compiler, Runtime and Tools.
Neuron SDK Release - January 05, 2022
Neuron 1.16.3 is a minor release. This release includes performance enhancements and operator support in PyTorch Neuron
and minor bug fixes in Neuron Compiler.
Neuron SDK Release - December 15, 2021
Neuron 1.16.2 is a patch release. This release include performance enhancements and minor bug fixes in Neuron Compiler and PyTorch Neuron.
Neuron SDK Release - November 5, 2021
Neuron 1.16.1 is a patch release. This release fixes a bug in Neuron Runtime that would have prevented users from launching a container that doesn’t use all of the Neuron Devices in the instance. If you are using Neuron within a container, please update to this new release by updating to latest Neuron ML framework package, Neuron Tools, and/or TensorFlow Neuron Model Server.
Neuron SDK Release - October 27, 2021
Neuron 1.16.0 is a release that requires your attention. You must update to the latest Neuron Driver ( aws-neuron-dkms version 2.1 or newer) for successful installation or upgrade.
This release introduces Neuron Runtime 2.x, upgrades PyTorch Neuron to PyTorch 1.9.1, adds support for new APIs (torch.neuron.DataParallel() and torch_neuron.is_available()), adds new features and capabilities (compiler --fast-math option for better fine-tuning of accuracy/performance and MXNet FlexEG feature), improves tools, adds support for additional operators, improves performance (Up to 20% additional throughput and up to 25% lower latency), and reduces model loading times. It also simplifies Neuron installation steps, and improves the user experience of container creation and deployment. In addition it includes bug fixes, new application notes, updated tutorials, and announcements of software deprecation and maintenance.
Neuron SDK Release - September 22, 2021
Neuron 1.15.2 includes bug fixes for the tensorflow-model-server-neuron package.
Neuron SDK Release - August 30, 2021
Neuron 1.15.1 includes bug fixes for the aws-neuron-dkms package and several other bug fixes for related packages.
Neuron SDK Release - August 12, 2021
Neuron 1.15.0 is the first release to support TensorFlow 2. In this release TensorFlow 2 supports language transformer base models like BERT. The TensorFlow 2 support will be enhanced in future releases to support additional models.
Neuron SDK Release - July 26, 2021
This release (Neuron 1.14.2) , include bug fixes and minor enhancements to Neuron Runtime:
Neuron Runtime - see Neuron Runtime Release Notes