TensorFlow r2.8 branch cut, December 16

Hi - we would like to announce that a new TF release 2.8.0 is coming up and we will be cutting the r2.8 branch on 12/16/2021 end of day. You can refer to TF 2.8.0 GitHub RELEASE.md document for the current plans for what’s upcoming, and will be updated with changes as we prepare the final release.

While this is mostly a maintenance release, a key focus continues to be on Performance, with several improvements planned in coordination with community contributors. We look forward to delivering these improvements in the final release in mid January 2022.

5 Likes

At the same time as the TF 2.8 release we will release patches for TF 2.7, 2.6 and 2.5.

1 Like

We met a weird bug on Colab by using 2.5, 2.6,2.7. Tensorflow code worked for GPU settings but using TPUs it crashed.

@Charlie_Li thank you for reporting this, would you be able to open an issue for this on GitHub: Issues · tensorflow/tensorflow · GitHub

We are delaying the branch cut as Windows support for Python 3.10 has encountered some issues.

We hope to be able to cut the branch early next week (Monday/Tuesday)

1 Like

Branch cut happened at commit a4e401da71458d253b05e41f28637b65baf64be4

3 Likes

Tensorflow 2.8 RC0 has been published. You can get it with pip install tensorflow==2.8.0rc0 (and similar commands)

Tensorflow 2.8 RC1 has been published. You can get it with pip install tensorflow==2.8.0rc1 (and similar commands)

Final release should arrive in ~1 week, pending no major regressions being found.

TensorFlow 2.8 final release has been published, as well as patches for versions from 2.5 to 2.7 inclusive.

See release notes for everything included in the release

2 Likes