Uh oh!
There was an error while loading. Please reload this page.
CI: Add Windows GPU runner for tests - #444
Conversation
leofang
commented
Feb 11, 2025
/ok to test |
This comment has been minimized.
This comment has been minimized.
leofang
commented
Feb 11, 2025
/ok to test |
leofang
commented
Feb 11, 2025
@cryos@rwgk do you guys have clever ways about refactoring a workflow file? On the Windows runner, we cannot launch any container so this part must be removed: cuda-python/.github/workflows/test-wheel.yml Lines 32 to 38 in 7387715 but otherwise the remaining logic should be the same and preserved... Are there smarter ways than duplicating the file for Windows? |
cryos
commented
Feb 11, 2025
I feel like there should be, searching turned up the same disappointing you cannot make |
leofang
commented
Feb 11, 2025
/ok to test |
1 similar comment
leofang
commented
Feb 11, 2025
/ok to test |
leofang
commented
Feb 11, 2025
/ok to test |
leofang
commented
Feb 11, 2025
(spoke to George, the underlying image is changed now, let me restart the tests) |
leofang
commented
Feb 11, 2025
/ok to test |
leofang
commented
Feb 11, 2025
/ok to test |
leofang
commented
Feb 12, 2025
Looks like we hit https://github.com/orgs/community/discussions/138813... which is weird... I haven't seen this error in the previous project... |
leofang
commented
Feb 14, 2025
Unfortunately there is no way to route this issue to the right GitHub support person. Closing this for now. |
leofang
commented
Feb 19, 2025
/ok to test |
leofang
commented
Feb 19, 2025
/ok to test |
leofang
commented
Feb 19, 2025
/ok to test |
leofang
commented
Feb 19, 2025
/ok to test |
leofang
commented
Feb 19, 2025
/ok to test |
leofang
commented
Feb 19, 2025
/ok to test |
leofang
commented
Feb 19, 2025
This is ready for review! The PR description is updated. |
leofang
commented
Feb 19, 2025
The doc build failure is due to Sphinx 8.2.0, let's fix it in a separate PR: |
Uh oh!
There was an error while loading. Please reload this page.
cryos
left a comment
There was a problem hiding this comment.
Very cool, looks good to me.
|
Part of #277.
Based on offline discussion with our admin @aflat and @aterrel@kkraus14, as a stop-gap solution until we get our own runners we are setting up a GH-hosted runners to cover Windows. Because the rate is high, we pick only 1 Python version x 2 latest CUDA major versions (11.8 & 12.8) x 2 different ways of getting CUDA to keep the cost within $100 a month.
In terms of the implementation, the new workflow is largely the same as its Linux counterpart, except for the following changes since the VM image of choice (Windows 11 23H2) has nothing pre-installed:
fetch_ctkaction cannot be used; instead, we switch to useJimver/cuda-toolkitwhich has been approved by the admins