sync with upstream #23
Merged
Conversation
This file contains 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
PR does a small change for non project FLASH path of the ci/gpu/build.sh script. We were observing a linking bug in the `test/ml` executable _only_ when it was being executed in the Ubuntu 18.04 containers (see https://gpuci.gpuopenanalytics.com/view/RAPIDS%20-%20GPU%20Matrix/job/rapidsai/job/gpuci/job/cuml/job/branches/job/cuml-gpu-matrix-branch-0.18/) This PR works around the bug for that particular situation by letting LD_LIBRARY_PATH contain the conda lib folder up to the point of execution of `test/ml`, and does not affect regular project FLASH PR CI. Authors: - Dante Gama Dessavre (@dantegd) Approvers: - Ray Douglass (@raydouglass) URL: #3495
Issues and PRs without activity for 30d will be marked as stale. If there is no activity for 90d, they will be marked as rotten. Authors: - Jordan Jacobelli (@Ethyling) Approvers: - Dillon Cullinan (@dillon-cullinan) URL: #3500
This PR adds the GitHub action [PR Labeler](https://github.com/actions/labeler) to auto-label PRs based on their content. Labeling is managed with a configuration file `.github/labeler.yml` using the following [options](https://github.com/actions/labeler#usage). Authors: - Joseph (@jolorunyomi) Approvers: - Mike Wendt (@mike-wendt) - Dante Gama Dessavre (@dantegd) - John Zedlewski (@JohnZed) URL: #3407
Follows #3500 Updates the stale GHA with the following changes: - [x] Uses `inactive-30d` and `inactive-90d` labels instead of `stale` and `rotten` - [x] Updates comments to reflect changes in labels - [x] Exempts the following labels from being marked `inactive-30d` or `inactive-90d` - `0 - Blocked` - `0 - Backlog` - `good first issue` Authors: - Mike Wendt (@mike-wendt) Approvers: - Ray Douglass (@raydouglass) URL: #3507
Fix merge conflicts in 3502 [skip ci]
closes #2546 This PR improves the warning message printed when max iterations are reached during fitting a linear model. Example: ```python import numpy as np from cuml.linear_model import LogisticRegression from sklearn.datasets import load_breast_cancer X, y = load_breast_cancer(return_X_y=True) y = y.astype(np.float64) cls = LogisticRegression(penalty='none', C=1) cls.fit(X, y) ``` This produces the following output, where the last line is added by this PR: ``` [W] [15:31:04.467478] L-BFGS: max iterations reached [W] [15:31:04.467804] Maximum iterations reached before solver is converged. To increase model accuracy you can increase the number of iterations (max_iter) or improve the scaling of the input data. ``` Authors: - Tamas Bela Feher (@tfeher) Approvers: - Dante Gama Dessavre (@dantegd) URL: #3515
Implements hinge_loss with tests. API matches and code is adapted from sklearn https://scikit-learn.org/stable/modules/generated/sklearn.metrics.hinge_loss.html NOTE: We're not using the C++ hinge implementation in order to comply with sklearn API. The C++ implementation is used in the intermediate step in SGD and accept different set of parameters than what is used to calculate hinge_loss. Authors: - Nanthini Balasubramanian (@Nanthini10) Approvers: - William Hicks (@wphicks) - Michael Demoret (@mdemoret-nv) URL: #3409
- Add ccache env variables to conda recipe - Use Ninja in CI Authors: - Jordan Jacobelli (@Ethyling) Approvers: - Dillon Cullinan (@dillon-cullinan) - Dante Gama Dessavre (@dantegd) URL: #3508
Closes #3377. Authors: - Micka (@lowener) Approvers: - Victor Lafargue (@viclafargue) - John Zedlewski (@JohnZed) URL: #3460
…#3551) Similar to #3550 but targetting branch-0.18, can remove [skip-ci] if/when we decide to merge it to 0.18 as opposed to 0.19 Authors: - @dantegd Approvers: - @mike-wendt - @raydouglass URL: #3551
…#3550) Auto merger PR that got lost: #3036 Authors: - Dante Gama Dessavre (@dantegd) Approvers: - AJ Schmidt (@ajschmidt8) - @jakirkham URL: #3550
Testing the following undocumented option: ```yml codecov: allow_coverage_offsets: true ``` Suggested by Codecov.io support [here](https://community.codecov.io/t/unable-to-determine-a-parent-commit-to-compare-against-in-base-branch-after-squash-and-merge/2480/15?u=mdemoret-nv) to fix the linked issue with missing base reports. Authors: - Michael Demoret (@mdemoret-nv) Approvers: - AJ Schmidt (@ajschmidt8) - John Zedlewski (@JohnZed) URL: #3524
From suggestion and talks with @mdemoret-nv we came to the conclusion that the `_get_tags` function is very useful as a class/static method, so this PR changes it. If we require tags that are instance based, we can extend/modify the method in the future, but for now this seems like the best balance of simplicity and usefulness. Authors: - Dante Gama Dessavre (@dantegd) Approvers: - Michael Demoret (@mdemoret-nv) URL: #3257
Fix merge conflicts in 3552 [skip ci]
Introduce a global_settings object implemented as a thread-local borg for storing global settings in a thread-safe way Ensure that global_output_type is manipulated only in thread-safe ways and provide associated unit tests Move "root" context manager to global_settings object to avoid unnecessary existence check for root_cm Close #3234 Mitigate #3237 Authors: - William Hicks (@wphicks) Approvers: - Michael Demoret (@mdemoret-nv) - Dante Gama Dessavre (@dantegd) URL: #3497
Follow-up to #3453 to expose the file excluder through command-line arguments. Authors: - Micka (@lowener) Approvers: - Michael Demoret (@mdemoret-nv) - AJ Schmidt (@ajschmidt8) - Dante Gama Dessavre (@dantegd) URL: #3466
* This PR adds NVTX Markers to major time-consuming function calls of the regressors and classifiers of RF and DecisionTrees. * They span both RandomForest and DecisionTree code-bases Authors: - Venkat (@venkywonka) - John Zedlewski (@JohnZed) Approvers: - Thejaswi. N. S (@teju85) - AJ Schmidt (@ajschmidt8) - John Zedlewski (@JohnZed) URL: #3014
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
No description provided.