tensorflow / recommenders-addons

Additional utils and helpers to extend TensorFlow when build recommendation systems, contributed and maintained by SIG Recommenders.

Apache License 2.0

592 stars 133 forks source link

[Feat] Compatible with TensorFlow 2.15 #386

Closed MoFHeka closed 8 months ago

MoFHeka commented 8 months ago

Description

Brief Description of the PR:

Be compatible with TensorFlow 2.15 by calling the origin python function from new directory.
Fix some test should not be run when eager mode. Also add more new test for Redis KV backend.
Fix Redis backend json config file would not printed when a wrong json value type was set.
Fix check wrong array index in kvs_reply from Redis returning. Wanted to check the length of value, but the length of key was checked.

Type of change

[x] Bug fix
[ ] New Tutorial
[ ] Updated or additional documentation
[x] Additional Testing
[ ] New Feature

Checklist:

[x] I've properly formatted my code according to the guidelines
- [x] By running yapf
- [x] By running clang-format
[ ] This PR addresses an already submitted issue for TensorFlow Recommenders-Addons
[x] I have made corresponding changes to the documentation
[x] I have added tests that prove my fix is effective or that my feature works

How Has This Been Tested?

Run all test script under TF2.15 environment.

rhdong commented 8 months ago

Hi @MoFHeka, we need a separate PR for switching the TF version in the CI and release.

jq commented 8 months ago

👍🏻 when will we have this release to https://pypi.org/project/tensorflow-recommenders-addons/0.6.0/ ?