Cuml Versions Save

cuML - RAPIDS Machine Learning Library

v24.02.00

1 month ago

🚨 Breaking Changes

  • Update to CCCL 2.2.0. (#5702) @bdice
  • Switch to scikit-build-core (#5693) @vyasr

🐛 Bug Fixes

  • [Hotfix] Fix FIL gtest (#5755) @hcho3
  • Exclude tests from builds (#5754) @vyasr
  • Fix ctest directory to ensure tests are executed (#5753) @bdice
  • Synchronize stream in SVC memory test (#5729) @wphicks
  • Fix shared-workflows repo name (#5723) @raydouglass
  • Fix cupy dependency in pyproject.toml (#5705) @vyasr
  • Only cufft offers a static_nocallback version of the library (#5703) @robertmaynard

🛠️ Improvements

  • [Hotfix] Update GPUTreeSHAP to fix ARM build (#5747) @hcho3
  • Disable HistGradientBoosting support for now (#5744) @hcho3
  • Disable hnswlib feature in RAFT; pin pytest (#5733) @hcho3
  • [LogisticRegressionMG] Support standardization with no data modification (#5724) @lijinf2
  • Remove usages of rapids-env-update (#5716) @KyleFromNVIDIA
  • Remove extraneous SKBUILD_BUILD_OPTIONS (#5714) @vyasr
  • refactor CUDA versions in dependencies.yaml (#5712) @jameslamb
  • Update to CCCL 2.2.0. (#5702) @bdice
  • Migrate to Treelite 4.0 (#5701) @hcho3
  • Use cuda::proclaim_return_type on device lambdas. (#5696) @bdice
  • move _process_generic to base_return_types, avoid circular import (#5695) @dcolinmorgan
  • Switch to scikit-build-core (#5693) @vyasr
  • Fix all deprecated function calls in TUs where warnings are errors (#5692) @vyasr
  • Remove CUML_BUILD_WHEELS and standardize Python builds (#5689) @vyasr
  • Forward-merge branch-23.12 to branch-24.02 (#5657) @bdice
  • Add cuML devcontainers (#5568) @trxcllnt

v24.04.00a

1 month ago

🐛 Bug Fixes

  • Update pre-commit-hooks to v0.0.3 (#5816) @KyleFromNVIDIA
  • Remove use of cudf.core.column.full. (#5794) @bdice
  • Suppress all HealthChecks on test_split_datasets. (#5791) @bdice
  • Suppress a hypothesis HealthCheck on test_split_datasets that fails in nightly CI. (#5790) @bdice
  • [BUG] Fix MAX_THREADS_PER_SM on sm 89. (#5785) @trivialfis
  • fix device to host copy not sync stream in logistic regression mg (#5766) @lijinf2
  • Use cudf.Index instead of cudf.GenericIndex. (#5738) @bdice
  • update RAPIDS dependencies to 24.4, refactor dependencies.yaml (#5726) @jameslamb

🚀 New Features

  • Support CUDA 12.2 (#5711) @jameslamb

🛠️ Improvements

  • Update script input name (#5802) @AyodeAwe
  • Add upper bound to prevent usage of NumPy 2 (#5797) @bdice
  • Enable pytest failures on warnings from cudf (#5796) @mroeschke
  • Use public cudf APIs where possible (#5795) @mroeschke
  • Remove hard-coding of RAPIDS version where possible (#5793) @KyleFromNVIDIA
  • Switch pytest-xdist algorithm to worksteal (#5792) @bdice
  • Automate C++ include file grouping and ordering using clang-format (#5787) @harrism
  • Add support for Python 3.11, require NumPy 1.23+ (#5786) @jameslamb
  • [ENH] Let cuDF handle input types for label encoder. (#5783) @trivialfis
  • Install test dependencies at the same time as cuml packages. (#5781) @bdice
  • Update devcontainers to CUDA Toolkit 12.2 (#5778) @trxcllnt
  • target branch-24.04 for GitHub Actions workflows (#5776) @jameslamb
  • Add environment-agnostic scripts for running ctests and pytests (#5761) @trxcllnt
  • Pandas 2.x support (#5758) @dantegd
  • Update ops-bot.yaml (#5752) @AyodeAwe
  • Forward-merge branch-24.02 to branch-24.04 (#5735) @bdice
  • Replace local copyright check with pre-commit-hooks verify-copyright (#5732) @KyleFromNVIDIA
  • DBSCAN utilize rbc eps_neighbors (#5728) @mfoerste4

v23.12.00

3 months ago

🚨 Breaking Changes

  • [LogisticRegressionMG] Support sparse vectors (#5632) @lijinf2

🐛 Bug Fixes

  • Update actions/labeler to v4 (#5686) @raydouglass
  • updated docs around make_column_transformer change from .preprocessing to .compose (#5680) @taureandyernv
  • Skip dask pytest NN hang in CUDA 11.4 CI (#5665) @dantegd
  • Avoid hard import of sklearn in base module. (#5663) @csadorf
  • CI: Pin clang-tidy to 15.0.7. (#5661) @csadorf
  • Adjust assumption regarding valid cudf.Series dimensional input. (#5654) @csadorf
  • Flatten cupy array before feeding to cudf.Series (#5651) @vyasr
  • CI: Fix expected ValueError and dask-glm incompatibility (#5644) @csadorf
  • Use drop_duplicates instead of unique for cudf's pandas compatibility mode (#5639) @vyasr
  • Temporarily avoid pydata-sphinx-theme version 0.14.2. (#5629) @csadorf
  • Fix type hint in split function. (#5625) @trivialfis
  • Fix trying to get pointer to None in svm/linear.pyx (#5615) @yosider
  • Reduce parallelism to avoid OOMs in wheel tests (#5611) @vyasr

📖 Documentation

  • Update interoperability docs (#5633) @beckernick
  • Update instructions for creating a conda build environment (#5628) @csadorf

🚀 New Features

  • Basic implementation of OrdinalEncoder. (#5646) @trivialfis

🛠️ Improvements

  • Build concurrency for nightly and merge triggers (#5658) @bdice
  • [LogisticRegressionMG][FEA] Support training when dataset contains only one class (#5655) @lijinf2
  • Use new rapids-dask-dependency metapackage for managing dask versions (#5649) @galipremsagar
  • Simplify some logic in LabelEncoder (#5648) @vyasr
  • Increase Nanny close timeout in LocalCUDACluster tests (#5636) @pentschev
  • [LogisticRegressionMG] Support sparse vectors (#5632) @lijinf2
  • Add rich HTML representation to estimators (#5630) @betatim
  • Unpin dask and distributed for 23.12 development (#5627) @galipremsagar
  • Update shared-action-workflows references (#5621) @AyodeAwe
  • Use branch-23.12 workflows. (#5618) @bdice
  • Update rapids-cmake functions to non-deprecated signatures (#5616) @robertmaynard
  • Allow nightly dependencies and set up consistent nightly versions for conda and pip packages (#5607) @vyasr
  • Forward-merge branch-23.10 to branch-23.12 (#5596) @bdice
  • Build CUDA 12.0 ARM conda packages. (#5595) @bdice
  • Enable multiclass svm for sparse input (#5588) @mfoerste4

v24.02.00a

3 months ago

🚨 Breaking Changes

  • Update to CCCL 2.2.0. (#5702) @bdice
  • Switch to scikit-build-core (#5693) @vyasr

🐛 Bug Fixes

  • Synchronize stream in SVC memory test (#5729) @wphicks
  • Fix shared-workflows repo name (#5723) @raydouglass
  • Fix cupy dependency in pyproject.toml (#5705) @vyasr
  • Only cufft offers a static_nocallback version of the library (#5703) @robertmaynard

🛠️ Improvements

  • Remove usages of rapids-env-update (#5716) @KyleFromNVIDIA
  • Remove extraneous SKBUILD_BUILD_OPTIONS (#5714) @vyasr
  • refactor CUDA versions in dependencies.yaml (#5712) @jameslamb
  • Update to CCCL 2.2.0. (#5702) @bdice
  • Use cuda::proclaim_return_type on device lambdas. (#5696) @bdice
  • move _process_generic to base_return_types, avoid circular import (#5695) @dcolinmorgan
  • Switch to scikit-build-core (#5693) @vyasr
  • Fix all deprecated function calls in TUs where warnings are errors (#5692) @vyasr
  • Remove CUML_BUILD_WHEELS and standardize Python builds (#5689) @vyasr
  • Forward-merge branch-23.12 to branch-24.02 (#5657) @bdice
  • Add cuML devcontainers (#5568) @trxcllnt

v23.10.00

5 months ago

🚨 Breaking Changes

  • add sample_weight parameter to dbscan.fit (#5574) @mfoerste4
  • Update to Cython 3.0.0 (#5506) @vyasr

🐛 Bug Fixes

  • Fix accidental unsafe cupy import (#5613) @dantegd
  • Fixes for CPU package (#5599) @dantegd
  • Fixes for timeouts in tests (#5598) @dantegd

🚀 New Features

  • Enable cuml-cpu nightly (#5585) @dantegd
  • add sample_weight parameter to dbscan.fit (#5574) @mfoerste4

🛠️ Improvements

  • cuml-cpu notebook, docs and cluster models (#5597) @dantegd
  • Pin dask and distributed for 23.10 release (#5592) @galipremsagar
  • Add changes for early experimental support for dataframe interchange protocol API (#5591) @dantegd
  • [FEA] Support L1 regularization and ElasticNet in MNMG Dask LogisticRegression (#5587) @lijinf2
  • Update image names (#5586) @AyodeAwe
  • Update to clang 16.0.6. (#5583) @bdice
  • Upgrade to Treelite 3.9.1 (#5581) @hcho3
  • Update to doxygen 1.9.1. (#5580) @bdice
  • [REVIEW] Adding a few of datasets for benchmarking (#5573) @vinaydes
  • Allow cuML MNMG estimators to be serialized (#5571) @viclafargue
  • [FEA] Support multiple classes in multi-node-multi-gpu logistic regression, from C++, Cython, to Dask Python class (#5565) @lijinf2
  • Use copy-pr-bot (#5563) @ajschmidt8
  • Unblock CI for branch-23.10 (#5561) @csadorf
  • Fix CPU-only build for new FIL (#5559) @hcho3
  • [FEA] Support no regularization in MNMG LogisticRegression (#5558) @lijinf2
  • Unpin dask and distributed for 23.10 development (#5557) @galipremsagar
  • Branch 23.10 merge 23.08 (#5547) @vyasr
  • Use Python builtins to prep benchmark tmp_dir (#5537) @jakirkham
  • Branch 23.10 merge 23.08 (#5522) @vyasr
  • Update to Cython 3.0.0 (#5506) @vyasr

v23.12.00a

5 months ago

🐛 Bug Fixes

  • Avoid hard import of sklearn in base module. (#5663) @csadorf
  • CI: Pin clang-tidy to 15.0.7. (#5661) @csadorf
  • Adjust assumption regarding valid cudf.Series dimensional input. (#5654) @csadorf
  • Flatten cupy array before feeding to cudf.Series (#5651) @vyasr
  • CI: Fix expected ValueError and dask-glm incompatibility (#5644) @csadorf
  • Use drop_duplicates instead of unique for cudf's pandas compatibility mode (#5639) @vyasr
  • Temporarily avoid pydata-sphinx-theme version 0.14.2. (#5629) @csadorf
  • Fix type hint in split function. (#5625) @trivialfis
  • Fix trying to get pointer to None in svm/linear.pyx (#5615) @yosider
  • Reduce parallelism to avoid OOMs in wheel tests (#5611) @vyasr

📖 Documentation

  • Update interoperability docs (#5633) @beckernick
  • Update instructions for creating a conda build environment (#5628) @csadorf

🚀 New Features

  • Basic implementation of OrdinalEncoder. (#5646) @trivialfis

🛠️ Improvements

  • Build concurrency for nightly and merge triggers (#5658) @bdice
  • Use new rapids-dask-dependency metapackage for managing dask versions (#5649) @galipremsagar
  • Simplify some logic in LabelEncoder (#5648) @vyasr
  • Increase Nanny close timeout in LocalCUDACluster tests (#5636) @pentschev
  • Add rich HTML representation to estimators (#5630) @betatim
  • Unpin dask and distributed for 23.12 development (#5627) @galipremsagar
  • Update shared-action-workflows references (#5621) @AyodeAwe
  • Use branch-23.12 workflows. (#5618) @bdice
  • Update rapids-cmake functions to non-deprecated signatures (#5616) @robertmaynard
  • Allow nightly dependencies and set up consistent nightly versions for conda and pip packages (#5607) @vyasr
  • Forward-merge branch-23.10 to branch-23.12 (#5596) @bdice
  • Build CUDA 12.0 ARM conda packages. (#5595) @bdice
  • Enable multiclass svm for sparse input (#5588) @mfoerste4

v23.08.00

7 months ago

🚨 Breaking Changes

  • Stop using setup.py in build.sh (#5500) @vyasr
  • Add copy_X parameter to LinearRegression (#5495) @viclafargue

🐛 Bug Fixes

  • Update dependencies.yaml test_notebooks to include dask_ml (#5545) @taureandyernv
  • Fix cython-lint issues. (#5536) @bdice
  • Skip rf_memleak tests (#5529) @dantegd
  • Pin hdbscan to fix pytests in CI (#5515) @dantegd
  • Fix UMAP and simplicial set functions metric (#5490) @viclafargue
  • Fix test_masked_column_mode (#5480) @viclafargue
  • Use fit_predict rather than fit for KNeighborsClassifier and KNeighborsRegressor in benchmark utility (#5460) @beckernick
  • Modify HDBSCAN membership_vector batch_size check (#5455) @tarang-jain

🚀 New Features

  • Use rapids-cmake testing to run tests in parallel (#5487) @robertmaynard
  • [FEA] Update MST Reduction Op (#5386) @tarang-jain
  • cuml: Build CUDA 12 packages (#5318) @vyasr
  • CI: Add custom GitHub Actions job to run clang-tidy (#5235) @csadorf

🛠️ Improvements

  • Pin dask and distributed for 23.08 release (#5541) @galipremsagar
  • Remove Dockerfile. (#5534) @bdice
  • Improve temporary directory handling in cuML (#5527) @jakirkham
  • Support init arguments in MNMG LogisticRegression (#5519) @lijinf2
  • Support predict in MNMG Logistic Regression (#5516) @lijinf2
  • Remove unused matrix.cuh and math.cuh headers to eliminate deprecation warnings. (#5513) @bdice
  • Update gputreeshap to use rapids-cmake. (#5512) @bdice
  • Remove raft specializations includes. (#5509) @bdice
  • Revert CUDA 12.0 CI workflows to branch-23.08. (#5508) @bdice
  • Enable wheels CI scripts to run locally (#5507) @divyegala
  • Default to nproc for PARALLEL_LEVEL in build.sh. (#5505) @csadorf
  • Fixed potential overflows in SVM, minor adjustments to nvtx ranges (#5504) @mfoerste4
  • Stop using setup.py in build.sh (#5500) @vyasr
  • Fix PCA test (#5498) @viclafargue
  • Update build dependencies (#5496) @csadorf
  • Add copy_X parameter to LinearRegression (#5495) @viclafargue
  • Sparse pca patch (#5493) @Intron7
  • Restrict HDBSCAN metric options to L2 #5415 (#5492) @Rvch7
  • Fix typos. (#5481) @bdice
  • Add multi-node-multi-gpu Logistic Regression in C++ (#5477) @lijinf2
  • Add missing stream argument to cub calls in workingset (#5476) @mfoerste4
  • Update to CMake 3.26.4 (#5464) @vyasr
  • use rapids-upload-docs script (#5457) @AyodeAwe
  • Unpin dask and distributed for development (#5452) @galipremsagar
  • Remove documentation build scripts for Jenkins (#5450) @ajschmidt8
  • Fix update version and pinnings for 23.08. (#5440) @bdice
  • Add cython-lint configuration. (#5439) @bdice
  • Unpin scikit-build upper bound (#5438) @vyasr
  • Fix some deprecation warnings in tests. (#5436) @bdice
  • Update raft::sparse::distance::pairwise_distance to new API (#5428) @divyegala

v23.10.00a

7 months ago

🚨 Breaking Changes

  • Update to Cython 3.0.0 (#5506) @vyasr

🛠️ Improvements

  • Use copy-pr-bot (#5563) @ajschmidt8
  • Unpin dask and distributed for 23.10 development (#5557) @galipremsagar
  • Branch 23.10 merge 23.08 (#5547) @vyasr
  • Use Python builtins to prep benchmark tmp_dir (#5537) @jakirkham
  • Branch 23.10 merge 23.08 (#5522) @vyasr
  • Update to Cython 3.0.0 (#5506) @vyasr

v23.06.00

9 months ago

🚨 Breaking Changes

  • Dropping Python 3.8 (#5385) @divyegala
  • Support sparse input for SVC and SVR (#5273) @mfoerste4

🐛 Bug Fixes

  • Fix for umap-pytest issue in Rocky8 and wheels in GHA nightly tests (#5458) @dantegd
  • Fixes for nightly GHA runs (#5446) @dantegd
  • Add missing RAFT cusolver_macros import and changes for recent cuDF updates (#5434) @dantegd
  • Fix kmeans pytest to correctly compute fp output error (#5426) @mdoijade
  • Add missing raft/matrix/matrix.cuh include (#5411) @benfred
  • Fix path to cumlprims_mg in build workflow (#5406) @divyegala
  • Fix path to cumlprims in build workflow (#5405) @vyasr
  • Pin to scikit-build<17.2 (#5400) @vyasr
  • Fix forward merge #5383 (#5384) @dantegd
  • Correct buffer move assignment in experimental FIL (#5372) @wphicks
  • Avoid invalid memory access in experimental FIL for large output size (#5365) @wphicks
  • Fix forward merge #5336 (#5345) @dantegd

📖 Documentation

  • Fix HDBSCAN docs and add membership_vector to cuml.cluster.hdbscan namespace (#5378) @beckernick
  • Small doc fix (#5375) @tarang-jain

🚀 New Features

  • Provide method for auto-optimization of FIL parameters (#5368) @wphicks

🛠️ Improvements

  • Fix documentation source code links (#5449) @ajschmidt8
  • Drop seaborn dependency. (#5437) @bdice
  • Make all nvtx usage go through safe imports (#5424) @dantegd
  • run docs nightly too (#5423) @AyodeAwe
  • Switch back to using primary shared-action-workflows branch (#5420) @vyasr
  • Add librmm to libcuml dependencies. (#5410) @bdice
  • Update recipes to GTest version >=1.13.0 (#5408) @bdice
  • Remove cudf from libcuml meta.yaml (#5407) @divyegala
  • Support CUDA 12.0 for pip wheels (#5404) @divyegala
  • Support for gtest 1.11+ changes (#5403) @dantegd
  • Update cupy dependency (#5401) @vyasr
  • Build wheels using new single image workflow (#5394) @vyasr
  • Revert shared-action-workflows pin (#5391) @divyegala
  • Fix logic for concatenating Treelite objects (#5387) @hcho3
  • Dropping Python 3.8 (#5385) @divyegala
  • Remove usage of rapids-get-rapids-version-from-git (#5379) @jjacobelli
  • [ENH] Add missing includes of rmm/mr/device/per_device_resource.hpp (#5369) @ahendriksen
  • Remove wheel pytest verbosity (#5367) @sevagh
  • support parameter 'class_weight' and method 'decision_function' in LinearSVC (#5364) @mfoerste4
  • Update clang-format to 16.0.1. (#5361) @bdice
  • Implement apply() in FIL (#5358) @hcho3
  • Use ARC V2 self-hosted runners for GPU jobs (#5356) @jjacobelli
  • Try running silhouette test (#5353) @vyasr
  • Remove uses-setup-env-vars (#5344) @vyasr
  • Resolve auto-merger conflicts between branch-23.04 & branch-23.06 (#5340) @galipremsagar
  • Solve merge conflict of PR #5327 (#5329) @dantegd
  • Branch 23.06 merge 23.04 (#5315) @vyasr
  • Support sparse input for SVC and SVR (#5273) @mfoerste4
  • Delete outdated versions.json. (#5229) @bdice

v23.08.00a

9 months ago

🚨 Breaking Changes

  • Stop using setup.py in build.sh (#5500) @vyasr

🐛 Bug Fixes

  • Fix test_masked_column_mode (#5480) @viclafargue
  • Use fit_predict rather than fit for KNeighborsClassifier and KNeighborsRegressor in benchmark utility (#5460) @beckernick
  • Modify HDBSCAN membership_vector batch_size check (#5455) @tarang-jain

🚀 New Features

  • Use rapids-cmake testing to run tests in parallel (#5487) @robertmaynard
  • cuml: Build CUDA 12 packages (#5318) @vyasr
  • CI: Add custom GitHub Actions job to run clang-tidy (#5235) @csadorf

🛠️ Improvements

  • Fixed potential overflows in SVM, minor adjustments to nvtx ranges (#5504) @mfoerste4
  • Stop using setup.py in build.sh (#5500) @vyasr
  • Fix PCA test (#5498) @viclafargue
  • Update build dependencies (#5496) @csadorf
  • Fix typos. (#5481) @bdice
  • Add missing stream argument to cub calls in workingset (#5476) @mfoerste4
  • Update to CMake 3.26.4 (#5464) @vyasr
  • use rapids-upload-docs script (#5457) @AyodeAwe
  • Unpin dask and distributed for development (#5452) @galipremsagar
  • Remove documentation build scripts for Jenkins (#5450) @ajschmidt8
  • Fix update version and pinnings for 23.08. (#5440) @bdice
  • Add cython-lint configuration. (#5439) @bdice
  • Unpin scikit-build upper bound (#5438) @vyasr
  • Fix some deprecation warnings in tests. (#5436) @bdice
  • Update raft::sparse::distance::pairwise_distance to new API (#5428) @divyegala