{"full_name":"lightgbm-org/LightGBM","owner":"lightgbm-org","name":"LightGBM","shard":"1","url":"https://github.com/lightgbm-org/LightGBM","description":"A fast, distributed, high performance gradient boosting (GBT, GBDT, GBRT, GBM or MART) framework based on decision tree algorithms, used for ranking, classification and many other machine learning tasks.","homepage":"https://lightgbm.readthedocs.io/en/latest/","topics":["gbdt","gbm","machine-learning","data-mining","distributed","lightgbm","gbrt","decision-trees","gradient-boosting","python","r","parallel","kaggle"],"tag_pages":["data-mining","decision-trees","distributed","gradient-boosting","kaggle","machine-learning","parallel","python","r"],"language":"C++","license":"MIT","avatar":"https://avatars.githubusercontent.com/u/266222920?v=4","created_at":"2016-08-05T05:45:50Z","pushed_at":"2026-11-12T04:47:04Z","metrics":{"stars":18983,"forks":4109,"watchers":416,"issues":3707,"prs":3802,"releases":35,"issues_open":447,"issues_closed":3260,"prs_open":95,"prs_closed":423,"prs_merged":3284,"discussions":0},"growth":{"day":{"stars":6,"forks":2,"base_day":"2026-11-15"}},"history":{"dates":["2026-11-10","2026-11-11","2026-11-12","2026-11-13","2026-11-14","2026-11-15","2026-11-16"],"stars":[18962,18966,18970,18973,18976,18977,18983],"forks":[4105,4104,4104,4105,4107,4107,4109],"watchers":[null,416,416,416,416,416,416],"issues_open":[null,447,447,447,447,447,447],"prs_open":[null,89,93,93,93,93,95]},"boards":{},"same_owner":[],"issues":[{"number":3679,"title":"Avoiding Exception \"Check failed: (best_split_info.right_count) > (0) at ...\" with a regression task","comments":56,"created_at":"2020-12-25T09:40:15Z"},{"number":2302,"title":"Feature Requests & Voting Hub","comments":49,"created_at":"2019-08-01T03:02:47Z"},{"number":1003,"title":"BUG in GPU histogram","comments":48,"created_at":"2017-10-20T19:03:23Z"},{"number":2793,"title":"lightgbm.basic.LightGBMError: Bug in GPU histogram! split 11937: 12, smaller_leaf: 10245, larger_leaf: 1704","comments":39,"created_at":"2020-02-21T16:58:05Z"},{"number":4074,"title":"Dask tests randomly fail with socket error code 54 or 104","comments":31,"created_at":"2021-03-15T23:05:07Z"},{"number":4946,"title":"Check failed: (best_split_info.left_count) > (0)","comments":30,"created_at":"2022-01-13T03:19:51Z"},{"number":5390,"title":"[ci] [dask] CI jobs failing with Dask 2022.7.1","comments":29,"created_at":"2022-07-26T21:37:18Z"},{"number":6599,"title":"[R-package] [gpu] Cmake version reported to be low when compiling the R package through build_r.R in Windows","comments":27,"created_at":"2024-08-09T18:11:13Z"}],"discussions":[],"latest_release":{"name":"v4.7.0","tag":"v4.7.0","published_at":"2026-07-18T20:20:12Z","body":"## Highlights\r\n\r\n**The Python package now supports `polars` inputs, thanks to `narwhals`.**\r\nThanks to @borchero for the main implementation and to everyone who contributed design work, especially on https://github.com/lightgbm-org/LightGBM/issues/6204 (@maxzw @detrin @jmoralez @lcrmorin @Ic3fr0g @MarcoGorelli @kylebarron @vyasr).\r\n\r\n**Significantly improved GPU support** including the first ROCm builds (https://github.com/lightgbm-org/LightGBM/pull/6086) and multi-GPU support for NVIDIA GPUs via NCCL (https://github.com/lightgbm-org/LightGBM/pull/6138). (@jeffdaily @shiyu1994 @StrikerRUS)\r\n\r\n**Improvements to distributed training / Dask**. Lots of bugfixes in this area, still ongoing work, thanks especially to @wagner-austin for all the help.\r\n\r\n**Some organizational changes**. The project has changed GitHub orgs (`Microsoft -> lightgbm-org`) and default branch names (`master -> main`).\r\n\r\n## Changes\r\n\r\n## 💡 New Features\r\n\r\n- [python-package] allow access to eval result attributes by name @jameslamb (#7161)\r\n- [python-package] Allow to pass Arrow array & Polars series as position @maxzw (#7260)\r\n- [c++][python-package] Leverage Arrow PyCapsule Interface for Arrow interoperability (including polars support) @borchero (#7275)\r\n- [python-package] publish nightlies to pypi.anaconda.org @jameslamb (#7245)\r\n- [python-package] Add decision_function() to LGBMClassifier @CoderWota (#7159)\r\n- [CUDA] Multi-GPU for CUDA Version @shiyu1994 (#6138)\r\n- [python-package] scikit-learn fit() methods: add eval_X, eval_y, deprecate eval_set @lorentzenchr (#6857)\r\n- [CUDA] Add CUDA version of binary error metric @nicklamiller (#7070)\r\n- [c++] Add CPU version of standard R-squared metric @nicklamiller (#7008)\r\n- [python-package] Follow symlinks to lib_lightgbm in library-loading @gmhewett (#6977)\r\n- [ROCm] add support for ROCm/HIP device @jeffdaily (#6086)\r\n- [python-package][sklearn] Support PyArrow Table as an input in scikit-learn methods @StrikerRUS (#6910)\r\n- [ci] [dask] test `lightgbm.dask` on macOS @jameslamb (#6677)\r\n\r\n## 🔨 Breaking\r\n\r\n- [ci] avoid overwriting installed 'build' in build-python.sh @jameslamb (#7356)\r\n- change default branch from 'master' to 'main' @jameslamb (#7287)\r\n- [CUDA] support CUDA 13, support user-supplied CUDA architecture list, update default GPU architectures @jameslamb (#7271)\r\n- [python-package] make some imports lazy: 'dask', 'graphviz', 'joblib', 'matplotlib', 'psutil' @jameslamb (#7340)\r\n- [python-package] Make Dataset.set_position() update self.position @maxzw (#7344)\r\n- [c++][python-package] Leverage Arrow PyCapsule Interface for Arrow interoperability (including polars support) @borchero (#7275)\r\n- [c++] Remove incorrect `const` modifier from Arrow-related C API @borchero (#7284)\r\n- [python-package] Bump minimum supported Python version to 3.10 @borchero (#7276)\r\n- [ci] [docs] update handling of artifacts, stop publishing zip archive of complete source code @jameslamb (#7080)\r\n- [c++] default to C++17 for most builds @jameslamb (#7016)\r\n- [ci] [C++] drop support for Visual Studio 2015 (fixes #6803) @jameslamb (#7022)\r\n- [R-package] [ci] drop support for R 3.x (fixes #7023) @jameslamb (#7038)\r\n- [CUDA] default to C++17 for CUDA builds @jameslamb (#7013)\r\n- [ci] Drop support for Python 3.7 and Python 3.8 @borchero (#6896)\r\n- [python-package] drop support for h2o datatable @jameslamb (#6894)\r\n- [python-package] Load parameters from model string @Samsagax (#6852)\r\n- [R-package] make finalize() in R6 classes private @jameslamb (#6833)\r\n\r\n## 🚀 Efficiency Improvement\r\n\r\n- [CUDA] Multi-GPU for CUDA Version @shiyu1994 (#6138)\r\n\r\n## 🐛 Bug Fixes\r\n\r\n- [c++] fail on incomplete binary file writes @wgu9 (#7348)\r\n- Fix calculation of percentiles for weighted training with regression objectives (including `mape`, `quantile`, and `regression_l1`) @JKDasondee (#7224)\r\n- [python-package] fix misleading feature name warning on sklearn 1.6+ predict() @FranciscoRMendes (#7232)\r\n- [python-package] Enable ruff rule `B905` @borcher","is_prerelease":false},"readme":true,"readme_zh":false,"news":[],"stale":false,"snapshot_at":"2026-11-16T09:23:14Z","history_start":"2026-09-06"}