head 1.10; access; symbols pkgsrc-2026Q2:1.10.0.4 pkgsrc-2026Q2-base:1.10 pkgsrc-2026Q1:1.10.0.2 pkgsrc-2026Q1-base:1.10 pkgsrc-2025Q4:1.7.0.8 pkgsrc-2025Q4-base:1.7 pkgsrc-2025Q3:1.7.0.6 pkgsrc-2025Q3-base:1.7 pkgsrc-2025Q2:1.7.0.4 pkgsrc-2025Q2-base:1.7 pkgsrc-2025Q1:1.7.0.2 pkgsrc-2025Q1-base:1.7 pkgsrc-2024Q4:1.5.0.4 pkgsrc-2024Q4-base:1.5 pkgsrc-2024Q3:1.5.0.2 pkgsrc-2024Q3-base:1.5 pkgsrc-2024Q2:1.4.0.4 pkgsrc-2024Q2-base:1.4 pkgsrc-2024Q1:1.4.0.2 pkgsrc-2024Q1-base:1.4 pkgsrc-2023Q4:1.2.0.6 pkgsrc-2023Q4-base:1.2 pkgsrc-2023Q3:1.2.0.4 pkgsrc-2023Q3-base:1.2 pkgsrc-2023Q2:1.2.0.2 pkgsrc-2023Q2-base:1.2; locks; strict; comment @# @; 1.10 date 2026.02.16.11.13.35; author adam; state Exp; branches; next 1.9; commitid p8XfBKIoau2AMAuG; 1.9 date 2026.01.10.11.39.10; author adam; state Exp; branches; next 1.8; commitid eYhcMrxmofP27QpG; 1.8 date 2025.12.26.16.36.54; author adam; state Exp; branches; next 1.7; commitid PL3reOimpTn5fWnG; 1.7 date 2025.02.07.07.06.23; author adam; state Exp; branches; next 1.6; commitid kHb2jXa5Wu55rvIF; 1.6 date 2025.01.06.11.37.27; author adam; state Exp; branches; next 1.5; commitid RYbr0Va0RycMXpEF; 1.5 date 2024.08.04.13.05.59; author adam; state Exp; branches; next 1.4; commitid 84xFZ42Rlex2rvkF; 1.4 date 2024.01.24.22.43.20; author adam; state Exp; branches; next 1.3; commitid hevgkvqcRkiUPKVE; 1.3 date 2024.01.19.14.36.17; author adam; state Exp; branches; next 1.2; commitid NievF4jfuWiEi4VE; 1.2 date 2023.06.19.08.03.48; author adam; state Exp; branches; next 1.1; commitid 2wYsDyaCEkZq0xtE; 1.1 date 2023.06.13.17.36.58; author adam; state Exp; branches; next ; commitid GONqMQ9maOe1nOsE; desc @@ 1.10 log @py-xgboost: updated to 3.2.0 3.2.0 We are excited to announce the XGBoost 3.2 release. This release features significant progress on multi-target tree support with vector leaf, enhanced GPU external memory training, various optimizations, and the removal of the deprecated CLI. @ text @$NetBSD: distinfo,v 1.9 2026/01/10 11:39:10 adam Exp $ BLAKE2s (xgboost-3.2.0.tar.gz) = 98a31bf52f8bc4918f0d2950ea3e995dcf27a7b23ab2d38c80b00d5f1aaccc38 SHA512 (xgboost-3.2.0.tar.gz) = a9cafe87b012dfbeea4c2937b02c593f3d3bd4708c6eb282796915725e2194df390fc0ea961b3124d47120dba829038591b84e157c2e45007b02df4d921e3295 Size (xgboost-3.2.0.tar.gz) = 1263936 bytes SHA1 (patch-cpp__src_dmlc-core_CMakeLists.txt) = d7365655159fd310c73fa1abbbeb86db88c6f671 SHA1 (patch-cpp__src_include_xgboost_collective_socket.h) = f6b0cca135d6d8f6086213d6c221db466592287e SHA1 (patch-xgboost_libpath.py) = 39b2538729e5e4408de81534cd2b720ed112775a @ 1.9 log @py-xgboost: updated to 3.1.3 3.1.3 Scikit-learn 1.8 compatibility fix Add ARM CUDA wheels for PyPI. [R] Fix off-by-one bug: nrounds=0 resulted in 2 iterations [R] Fix mingw warnings, winbuilder check warnings, memory safety issues. Avoid overflow in rounding estimation. Workaround compiler issue on Windows, affects the use of max_delta_step with CUDA. @ text @d1 1 a1 1 $NetBSD: distinfo,v 1.8 2025/12/26 16:36:54 adam Exp $ d3 3 a5 3 BLAKE2s (xgboost-3.1.3.tar.gz) = f926b46a1473eef7df4e97c5f4471485885ac5f18ce32b9a9a8a2a5ec455a965 SHA512 (xgboost-3.1.3.tar.gz) = 0b1e3de74690a8c492634db1d8116611b60fcce0483f8a973f983a2dda41f5db20a21cb08de03f4bf93dd8b0ff3e949b79b887e4ec6ed61191c81159fc08d400 Size (xgboost-3.1.3.tar.gz) = 1237662 bytes @ 1.8 log @py-xgboost: updated to 3.1.2 3.1.2 Fix ordering of Python callbacks. Fix loading nccl 2.28. Infer the enable_categorical during model load. @ text @d1 1 a1 1 $NetBSD: distinfo,v 1.7 2025/02/07 07:06:23 adam Exp $ d3 3 a5 3 BLAKE2s (xgboost-3.1.2.tar.gz) = b3ec6a9c1b190d7238c487d381e88b5447fc3946e7b77c45f1f958c4c7b6b79a SHA512 (xgboost-3.1.2.tar.gz) = 78b1d3115b8fea73977c0428dfa9e3057777e2a4d8b6028b5eaf837ae64de5cc34625a6094cd97598063960ac262fd0ae307c8eb4e33feea4b0217841819a2e3 Size (xgboost-3.1.2.tar.gz) = 1237438 bytes @ 1.7 log @py-xgboost: updated to 2.1.4 2.1.4 The 2.1.4 patch release incorporates the following fixes on top of the 2.1.3 release: XGBoost is now compatible with scikit-learn 1.6 Build wheels with CUDA 12.8 and enable Blackwell support Adapt to RMM 25.02 logger changes @ text @d1 1 a1 1 $NetBSD: distinfo,v 1.6 2025/01/06 11:37:27 adam Exp $ d3 3 a5 3 BLAKE2s (xgboost-2.1.4.tar.gz) = 74970045fd7dfef89e52fdd72245fc1fd9a7bc1222b25d9eba07f6df0fee3bb6 SHA512 (xgboost-2.1.4.tar.gz) = 91385de1661daa449700f2cf09cf1276f302d14e0d9b77ee5dd441b840ba2defa5b2a280b30a315808d255c8cb39978947a4055f57c8ec1dee7e01229793a7cf Size (xgboost-2.1.4.tar.gz) = 1091127 bytes @ 1.6 log @py-xgboost: updated to 2.1.3 2.1.3 [pyspark] Support large model size Fix rng for the column sampler Handle cudf.pandas proxy objects properly 2.1.2 Clean up and modernize release-artifacts.py Fix ellpack categorical feature with missing values. Fix unbiased ltr with training continuation. Fix potential race in feature constraint. Fix boolean array for arrow-backed DF. Ensure that pip check does not fail due to a bad platform tag Check cub errors Limit the maximum number of threads. Fixes for large size clusters. POSIX compliant poll.h and mmap @ text @d1 1 a1 1 $NetBSD: distinfo,v 1.5 2024/08/04 13:05:59 adam Exp $ d3 3 a5 3 BLAKE2s (xgboost-2.1.3.tar.gz) = d0ac585d134b39c14159ddbc6987158de4f4c54e1101a3ebb5a145efa6f01d74 SHA512 (xgboost-2.1.3.tar.gz) = 7a53d2b47fe641724c62bb7cfb514e68eff983a452d2aa520258fcf2a6b7e926e30b723cbad03ecc6639d0673bec598faa0a69e14f48d5e518e02965fe166c43 Size (xgboost-2.1.3.tar.gz) = 1090326 bytes @ 1.5 log @py-xgboost: updated to 2.1.1 The 2.1.1 patch release make the following bug fixes: [Dask] Disable broadcast in the scatter call so that predict function won't hang [Dask] Handle empty partitions correctly Fix federated learning for the encrypted GRPC backend Fix a race condition in column splitter Gracefully handle cases where system files like /sys/fs/cgroup/cpu.max are not readable by the user Fix build and C++ tests for FreeBSD Clarify the requirement Pandas 1.2+ More robust endianness detection in R package build In addition, it contains several enhancements: Publish JVM packages targeting Linux ARM64 Publish a CPU-only wheel under name xgboost-cpu Support building with CUDA Toolkit 12.5 and latest CCCL @ text @d1 1 a1 1 $NetBSD: distinfo,v 1.4 2024/01/24 22:43:20 adam Exp $ d3 3 a5 3 BLAKE2s (xgboost-2.1.1.tar.gz) = 3f44b53d4371a3319a707551d67b522c8ee84c4a37ff24784a72d9a7dd815080 SHA512 (xgboost-2.1.1.tar.gz) = c3a0dc924aceaef21a3333558a207f0bf832f9550f254a73b11eadc1daac46574282ded3f034429f64ee2de9b4ab08337df463b497778e78b5fe76a5ada71866 Size (xgboost-2.1.1.tar.gz) = 1089959 bytes @ 1.4 log @py-xgboost: fix build @ text @d1 1 a1 1 $NetBSD: distinfo,v 1.3 2024/01/19 14:36:17 adam Exp $ d3 3 a5 3 BLAKE2s (xgboost-2.0.3.tar.gz) = 3f4382c6559f9d9cbbcb5299c9422fc1b194d3e9c2e27c456a9f1ed33eaf0bfa SHA512 (xgboost-2.0.3.tar.gz) = 93614a9ad9d0a256cc31586b701c46eef4353df76c3eac26f39df23c8c02fa9ec95e72a0cea0b51bc3e416b81b3ac557ed361afeda246376a7b561bb6f7da579 Size (xgboost-2.0.3.tar.gz) = 1048322 bytes d7 1 a7 1 SHA1 (patch-cpp__src_include_xgboost_collective_socket.h) = 71c2e47527fc39ca5cac4f47523f0f9ed8d22657 @ 1.3 log @py-xgboost: updated to 2.0.3 2.0.3 [backport][sklearn] Fix loading model attributes. [backport][py] Use the first found native library. [backport] [CI] Upload libxgboost4j.dylib (M1) to S3 bucket [jvm-packages] Fix POM for xgboost-jvm metapackage @ text @d1 1 a1 1 $NetBSD: distinfo,v 1.2 2023/06/19 08:03:48 adam Exp $ d6 1 a6 1 SHA1 (patch-cpp__src_dmlc-core_CMakeLists.txt) = 2b09910ae40fd26fdaf7462ad4b37662dd106554 @ 1.2 log @py-xgboost: updated to 1.7.6 1.7.6 Patch Release Bug Fixes Fix distributed training with mixed dense and sparse partitions. Fix monotone constraints on CPU with large trees. [spark] Make the spark model have the same UID as its estimator Optimize prediction with QuantileDMatrix. Document Improve doxygen Update the cuDF pip index URL. Maintenance Fix tests with pandas 2.0. @ text @d1 1 a1 1 $NetBSD: distinfo,v 1.1 2023/06/13 17:36:58 adam Exp $ d3 6 a8 6 BLAKE2s (xgboost-1.7.6.tar.gz) = 0802b62e76a5bd63dd4bc33c47d6f2da473b973a7e90196a86116b4e3b35a937 SHA512 (xgboost-1.7.6.tar.gz) = fe49bc72e1f28584c8cf739e3762da707f29c99f83123173619011d1d0f7600fb665d2e402e26873a7eb6b383a665fd2e96f5438dc15049227864e7583196831 Size (xgboost-1.7.6.tar.gz) = 848864 bytes SHA1 (patch-xgboost_dmlc-core_CMakeLists.txt) = 39432a69b84b177ab3bb882b960d9141f92771bd SHA1 (patch-xgboost_include_xgboost_collective_socket.h) = 0afa5538ae7b415af2e6901c995a84fa5539f86b SHA1 (patch-xgboost_libpath.py) = f9e2dbad7cf1831354f1174a939878b758eba57d @ 1.1 log @py-xgboost: added version 1.7.5 XGBoost is an optimized distributed gradient boosting library designed to be highly efficient, flexible and portable. It implements machine learning algorithms under the Gradient Boosting framework. XGBoost provides a parallel tree boosting (also known as GBDT, GBM) that solve many data science problems in a fast and accurate way. The same code runs on major distributed environment (Kubernetes, Hadoop, SGE, Dask, Spark, PySpark) and can solve problems beyond billions of examples. @ text @d1 1 a1 1 $NetBSD: distinfo,v 1.10 2023/05/08 08:51:03 adam Exp $ d3 3 a5 3 BLAKE2s (xgboost-1.7.5.tar.gz) = 2d15184249862c0f41a5d072d1274ad4d317f03d706a634296e7dd9d6a47cba5 SHA512 (xgboost-1.7.5.tar.gz) = 655060371060ba48da0675d24d41f3437bba937988dba1c388e73ab2db7642fd0e8b00ec5f6fa0d9d3ce20ac1c253ae798bc6034695449c8924373e6c84735c2 Size (xgboost-1.7.5.tar.gz) = 849023 bytes @