ray
ray copied to clipboard
Numpy 1.24 Support
Context
Numpy 1.24.0 was released on 12/18/22 and contains some breaking changes. This ticket tracks the changes needed to ensure Ray is compatible - full compatibility is targeted for the 2.3 release.
Note: This version was released for Python 3.8-3.11, so code/tests using an earlier version (i.e. 3.7) will not surface these incompatibilities.
Failing Tests
- https://github.com/ray-project/ray/issues/31193
- https://github.com/ray-project/ray/issues/31194
Mitigation
- https://github.com/ray-project/ray/pull/31240
Fixes
- https://github.com/ray-project/ray/issues/31182
- https://github.com/ray-project/ray/issues/31258
- https://github.com/ray-project/ray/issues/31277
I think the only remaining issues are #31182 and #31293.
- Why is there no test failing for this?
- Can we prioritize fixing that?
@krfricke I'm on-call this week and will focus on #31293
@matthewdeng I tested Numpy 1.24.2 with Ray 2.4.0 and it seems failed. Do we have a timeline to support this?
@YQ-Wang can you share what you're running and what the error you're getting is?
@matthewdeng Thanks for replying. I just want to confirm if 1.24.2 is fully supported because this ticket is not closed
@YQ-Wang yes ray should be fully compatible with numpy 1.24, which is the version contains in the Docker images. As always, it is still possible that other dependencies maybe introduce incompatibilities with numpy.
If you run into any issues, please re-open this issue!