Forward-merge release/26.04 into main - #7905
Merged
Merged
Conversation
This PR sets an upper bound on the `numba-cuda` dependency to `<0.29.0` Authors: - https://github.com/brandon-b-miller - Jim Crist-Harif (https://github.com/jcrist) Approvers: - Gil Forsyth (https://github.com/gforsyth) - Jim Crist-Harif (https://github.com/jcrist) URL: #7900
Contributor
Author
|
FAILURE - Unable to forward-merge due to an error, manual merge is necessary. Do not use the IMPORTANT: When merging this PR, do not use the auto-merger (i.e. the |
github-actions
Bot
added
conda
Closes #3148 Authors: - Simon Adorf (https://github.com/csadorf) Approvers: - Victor Lafargue (https://github.com/viclafargue) - Jim Crist-Harif (https://github.com/jcrist) - Gil Forsyth (https://github.com/gforsyth) URL: #7887
Addresses rapidsai/build-planning#247 Authors: - Simon Adorf (https://github.com/csadorf) - Jim Crist-Harif (https://github.com/jcrist) Approvers: - Divye Gala (https://github.com/divyegala) - Jim Crist-Harif (https://github.com/jcrist) - Gil Forsyth (https://github.com/gforsyth) URL: #7857
Member
|
Addressing conflicts in #7910. |
Closes #7827 The API reference section is now structured a bit more nicely. On `main` it is just one big page and it is hard to see what is what, what is where and who is who. 🖼️ Rendered preview https://downloads.rapids.ai/ci/cuml/pull-request/7798/55e9d3b/docs/cuml/html/api/ (can be outdated because it contains part of the commit hash) With this PR the API reference makes use of subsections for the various modules and maintains a list of "all the things" on that first API page (makes it easy to ctrl-f for a class you know exists and just want to get to fast): <img width="1265" height="955" alt="Screenshot 2026-02-13 at 11 56 31" src="https://github.com/user-attachments/assets/6d9d9288-0292-4e49-a7a4-d2ed4d9d42f1" /> You can navigate into a subsection and see the classes and functions available in it: <img width="1263" height="990" alt="Screenshot 2026-02-13 at 11 57 49" src="https://github.com/user-attachments/assets/2432b3e2-e156-4b7c-891d-82edcb452fb3" /> And you can visit a particular class/function to read everything there is to know about that class: <img width="1262" height="987" alt="Screenshot 2026-02-13 at 11 58 39" src="https://github.com/user-attachments/assets/a8e27fe7-1a19-41fe-9eaf-a2fb71ad66f4" /> --- There is an issue about restructuring the documentation, and we've discussed it a few times in the past. This morning I was frustrated enough while navigating the API reference section that I did this. I think it is a improvement on the status quo. It does surface that we could be more consistent with how we format our docstrings (eg without a linebreak after the first sentence all of the docstring ends up in the overview table or the fact that some metrics are shown directly on the "cuml.metrics" page but others referred to), etc but that is something we can tackle in a new issue I think. What do people think? I'm happy to do more polishing and think about HTTP redirects we need and such, but it would be good to know if people agree that this is going in the right direction. Authors: - Tim Head (https://github.com/betatim) - Simon Adorf (https://github.com/csadorf) Approvers: - Gil Forsyth (https://github.com/gforsyth) - Simon Adorf (https://github.com/csadorf) - Jim Crist-Harif (https://github.com/jcrist) URL: #7798
…vailable (#7916) We bumped the default devcontainer Python version to 3.14 in rapidsai/devcontainers#677 but some repos (like `cuml`) have optional dependencies without Python 3.14 support. We need to override the Python version temporarily until those deps catch up. I'll also open a tracking issue so we remember to undo this. Authors: - Gil Forsyth (https://github.com/gforsyth) Approvers: - Bradley Dice (https://github.com/bdice) URL: #7916
#7907) Contributes to rapidsai/build-planning#257 * builds CUDA 13 wheels with the 13.0 CTK * ensures wheels ship with a runtime dependency of `nvidia-nvjitlink>={whatever-minor-version-they-were-built-against}` Contributes to rapidsai/build-planning#256 * updates wheel tests to cover a range of CTK versions (we previously, accidentally, were only testing the latest 12.x and 13.x) ## Notes for Reviewers ### Doesn't this need conda changes too? I don't think so. cuML doesn't directly use libnvjitlink in conda packages, and it dynamically links to libcuvs: https://github.com/rapidsai/cuml/blob/95b7be311789ef6b66bf1983d09f3cfd17b5e508/cpp/CMakeLists.txt#L64 https://github.com/rapidsai/cuml/blob/main/conda/recipes/libcuml/recipe.yaml#L118-L126 Wheels need a runtime dependency on `nvidia-nvjitlink` because they statically link cuVS: https://github.com/rapidsai/cuml/blob/95b7be311789ef6b66bf1983d09f3cfd17b5e508/python/libcuml/CMakeLists.txt#L54 Authors: - James Lamb (https://github.com/jameslamb) Approvers: - Gil Forsyth (https://github.com/gforsyth) - Robert Maynard (https://github.com/robertmaynard) URL: #7907
After MNMG KMeans fit, the client was calling .result() on every worker's future, pulling the full fitted estimator (including `cluster_centers_`) from all workers back to the client. Since cluster centers are synchronized across workers via NCCL, every copy is identical making all but one transfer redundant. This PR changes the post-fit aggregation to: - Collect the full model from only the first worker - Extract `labels_` and `inertia_` from the remaining workers remotely via client.submit(getattr, ...) Authors: - Victor Lafargue (https://github.com/viclafargue) - Jim Crist-Harif (https://github.com/jcrist) Approvers: - Jim Crist-Harif (https://github.com/jcrist) URL: #7908
Numpy 2.4 no longer coerces 1 element arrays to scalars automatically. Small update to this line to support both 1-element alpha arrays and scalars. Fixes #7873. Authors: - Jim Crist-Harif (https://github.com/jcrist) Approvers: - Simon Adorf (https://github.com/csadorf) URL: #7918
Closes #7191 Authors: - Philip Hyunsu Cho (https://github.com/hcho3) Approvers: - Simon Adorf (https://github.com/csadorf) - Jim Crist-Harif (https://github.com/jcrist) - James Lamb (https://github.com/jameslamb) URL: #7858
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
10 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Forward-merge triggered by push to release/26.04 that creates a PR to keep main up-to-date. If this PR is unable to be immediately merged due to conflicts, it will remain open for the team to manually merge. See forward-merger docs for more info.