Skip to content

Conversation

jayfurmanek
Copy link

Update to 2.19.1

tensorflow-jenkins and others added 17 commits February 28, 2025 22:55
…umbers-2.19.0-30529

Update version numbers for TensorFlow 2.19.0
Update RELEASE.md with minor fix to indentation
r2.19 cherry-pick: 3a79ec9 "Bump libtpu versions to pick correct versioned nightlies"
* Fix saved_model.save for Serving embedding.

PiperOrigin-RevId: 740950472

* fix model reshard for sparsecore

---------

Co-authored-by: A. Unique TensorFlower <[email protected]>
Co-authored-by: Daniel Ng <[email protected]>
Update RELEASE.md with 2.19.1 release notes
…umbers-2.19.1-18073

Update version numbers for TensorFlow 2.19.1
Upgrade the h5py version in requirement_lock files.
PiperOrigin-RevId: 753622343
Copy link
Collaborator

@i-chaochen i-chaochen left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

As a side note, this UT is failed again, I thought we already fixed it in #3055 ? maybe we need to change it for all gfx... @AleksaArsic @ScXfjiang

[2025-09-05T23:08:25.409Z] //tensorflow/python/distribute:multi_process_runner_test_gpu             FAILED in 652.7s

@hsharsha
Copy link

I think the filtering of multi gpu test is not added into 2.19 branch?

@i-chaochen i-chaochen merged commit c214968 into r2.19-rocm-enhanced Sep 11, 2025
4 of 6 checks passed
@AleksaArsic
Copy link

AleksaArsic commented Sep 11, 2025

As a side note, this UT is failed again, I thought we already fixed it in #3055 ? maybe we need to change it for all gfx... @AleksaArsic @ScXfjiang

[2025-09-05T23:08:25.409Z] //tensorflow/python/distribute:multi_process_runner_test_gpu             FAILED in 652.7s

We have fixed the timeout in one of the tests from that particular UT suite - MultiProcessRunnerTest.test_global_pool. The issue could be in some other test in the UT suite. This needs further investigation.

CC @i-chaochen @ScXfjiang

@i-chaochen
Copy link
Collaborator

I think the filtering of multi gpu test is not added into 2.19 branch?

As a side note, this UT is failed again, I thought we already fixed it in #3055 ? maybe we need to change it for all gfx... @AleksaArsic @ScXfjiang

[2025-09-05T23:08:25.409Z] //tensorflow/python/distribute:multi_process_runner_test_gpu             FAILED in 652.7s

We have fixed the timeout in one of the tests from that particular UT suite - MultiProcessRunnerTest.test_global_pool. The issue could be in some other test in the UT suite. This needs further investigation.

CC @i-chaochen @ScXfjiang

@hsharsha could you have a look please?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

10 participants