We are encouraging users to update to at least TensorFlow 2.9.1 to ensure their installations do not receive undesired warnings. ________________________________
TensorFlow 2.9.1 Released
In response to reports of unexpected behavior, we have released TensorFlow 2.9.1 which is a patch release that fixes the CVE-2022-36002 vulnerability. This vulnerability was introduced in TensorFlow 2.8 and has been fixed in TensorFlow 2.9 via additional code reviews and RTL (runtime library) fixes.
Please update your installations with: pip install tensorflow==2.9.1
TensorFlow 2.9.1 Things to Know
This release fixes a critical vulnerability (CVE-2022-36002) in TensorFlow. Users who are running the latest version of TensorFlow 2.9.1 can avoid this warning by ensuring they are using at least the version 2.9.1 of TensorFlow, which was released on December 28th 2017.
TensorFlow is an open source software library for numerical computation used by both researchers and data scientists to build models and perform inference in deep neural networks as well as other machine learning models or applications that require numerical computations such as image processing and natural language processing.
The most recent release includes significant improvements to performance, usability, and stability, including:
• Significant improvements to performance in both CPU and GPU computations
• Significant improvements to usability
• Major updates to the APIs, including new functionality for users interested in building custom machine learning services
• New functionality to allow developers to deploy automated machine learning services
• Improved stability with more robust checkpointing
• Other minor bug fixes
Timeline
Published on: 09/16/2022 23:15:00 UTC
Last modified on: 09/20/2022 14:42:00 UTC