Files
lightgbm-org--lightgbm/README.md
T
wehub-resource-sync f3cad95168
Release Drafter / updateReleaseDraft (push) Has been cancelled
Build / archive (push) Has been cancelled
Python-package / bdist wheel (manylinux2014-arm, gcc, Python 3.14) (push) Has been cancelled
Python-package / sdist (windows-2022, MSVC, Python 3.10) (push) Has been cancelled
Python-package / sdist (macos-15-intel, clang, Python 3.10) (push) Has been cancelled
Python-package / mpi pip (ubuntu22.04, clang-17, Python 3.12) (push) Has been cancelled
Python-package / mpi source (ubuntu22.04, clang-17, Python 3.13) (push) Has been cancelled
Python-package / regular (ubuntu22.04, clang-17, Python 3.13) (push) Has been cancelled
Python-package / sdist (ubuntu22.04, clang-17, Python 3.14) (push) Has been cancelled
C++ / if-else (macos-15-intel, clang) (push) Has been cancelled
C++ / cpp-tests (macos-15-intel, clang) (push) Has been cancelled
C++ / if-else (ubuntu22.04, clang-17) (push) Has been cancelled
C++ / cpp-tests (ubuntu22.04, clang-17) (push) Has been cancelled
C++ / if-else (manylinux_2_28, gcc) (push) Has been cancelled
C++ / cpp-tests (windows-2022, msvc) (push) Has been cancelled
Python-package / gpu source (manylinux_2_28, gcc, Python 3.14) (push) Has been cancelled
Python-package / mpi source (macos-15-intel, gcc, Python 3.12) (push) Has been cancelled
Python-package / regular (manylinux_2_28, gcc, Python 3.11) (push) Has been cancelled
Python-package / bdist wheel (manylinux_2_28, gcc, Python 3.10) (push) Has been cancelled
Python-package / regular (macos-15-intel, clang, Python 3.11) (push) Has been cancelled
Python-package / bdist wheel (macos-15-intel, clang, Python 3.13) (push) Has been cancelled
Python-package / bdist wheel (macos-14, clang, Python 3.11) (push) Has been cancelled
Python-package / sdist (manylinux_2_28, gcc, Python 3.10) (push) Has been cancelled
Python-package / mpi wheel (ubuntu22.04, clang-17, Python 3.10) (push) Has been cancelled
Python-package / mpi source (manylinux_2_28, gcc, Python 3.11) (push) Has been cancelled
Python-package / regular (windows-2022, MSVC, Python 3.11) (push) Has been cancelled
Python-package / bdist wheel (windows-2022, MSVC, Python 3.13) (push) Has been cancelled
Python-package / Python - latest versions (manylinux_2_28) (push) Has been cancelled
Python-package / Python - oldest supported versions (manylinux_2_28, Python 3.10) (push) Has been cancelled
R-package / r-package (windows-2022, MSVC, R 4.3, cmake) (push) Has been cancelled
R-package / r-package (macos-15-intel, clang, R 4.3, cmake) (push) Has been cancelled
R-package / r-package (ubuntu22.04, clang, R 4.3, cmake) (push) Has been cancelled
R-package / r-package (ubuntu22.04, gcc, R 4.3, cmake) (push) Has been cancelled
R-package / r-package (windows-latest, MINGW, R 4.3, cran) (push) Has been cancelled
R-package / r-package (macos-14, clang, R 4.3, cran) (push) Has been cancelled
R-package / r-package (macos-15-intel, clang, R 4.3, cran) (push) Has been cancelled
R-package / r-package (clang21, R-devel) (push) Has been cancelled
R-package / r-package (windows-latest, MINGW, R 4.3, cmake) (push) Has been cancelled
R-package / r-package (ubuntu22.04, gcc, R 4.3, cran) (push) Has been cancelled
R-package / r-sanitizers (ubuntu-latest, R-devel, clang ASAN/UBSAN) (push) Has been cancelled
R-package / r-sanitizers (ubuntu-latest, R-devel, gcc ASAN/UBSAN) (push) Has been cancelled
R-package / r-package (clang22, R-devel) (push) Has been cancelled
R-package / r-package (gcc16, R-devel) (push) Has been cancelled
R-package / r-package (rchk, R-devel) (push) Has been cancelled
R-package / r-package (ubuntu-clang, R-devel) (push) Has been cancelled
R-package / r-package (ubuntu-gcc12, R-devel) (push) Has been cancelled
Static Analysis / lint (push) Has been cancelled
Static Analysis / check-docs (push) Has been cancelled
Static Analysis / r-package-check-docs (push) Has been cancelled
SWIG / swig (manylinux_2_28, gcc) (push) Has been cancelled
SWIG / swig (macos-15-intel, clang) (push) Has been cancelled
SWIG / swig (windows-2022, msvc) (push) Has been cancelled
SWIG / all-swig-jobs-successful (push) Has been cancelled
Build / all-build-jobs-successful (push) Has been cancelled
C++ / all-cpp-jobs-successful (push) Has been cancelled
Python-package / NuGet package (push) Has been cancelled
Python-package / upload-nightlies (push) Has been cancelled
Python-package / all-python-package-jobs-successful (push) Has been cancelled
R-package / all-r-package-jobs-successful (push) Has been cancelled
Static Analysis / all-static-analysis-jobs-successful (push) Has been cancelled
docs: make Chinese README the default
2026-07-13 10:38:42 +00:00

13 KiB
Raw Blame History

Note

本文档由 WeHub 基于上游 README 翻译整理,属于社区翻译,非官方中文文档。
English · 原始项目 · 上游 README
原作者、版权与许可证归属以原始项目及本仓库 LICENSE 文件为准。

Note

本项目已于 2026 年 3 月从 Microsoft/LightGBM 迁移至 lightgbm-org/LightGBM。 本仓库仍是 LightGBM 的官方源代码,由同一批维护者(包括 LightGBM 的创建者)管理。 详情请参阅 https://github.com/lightgbm-org/LightGBM/issues/7187

Light Gradient Boosting Machine

C++ GitHub Actions Build Status Python-package GitHub Actions Build Status R-package GitHub Actions Build Status CUDA Version GitHub Actions Build Status SWIG Wrapper GitHub Actions Build Status Static Analysis GitHub Actions Build Status Appveyor Build Status Documentation Status Link checks License EffVer Versioning StackOverflow questions Python Versions PyPI Version conda Version CRAN Version NuGet Version Winget Version

LightGBM 是一个采用基于树的学习算法的梯度提升(gradient boosting)框架。它面向分布式与高效场景设计,具有以下优势:

  • 更快的训练速度与更高的效率。
  • 更低的内存占用。
  • 更好的准确率。
  • 支持并行、分布式与 GPU 学习。
  • 能够处理大规模数据。

更多细节请参阅 Features.

得益于这些优势,LightGBM 已广泛应用于众多机器学习竞赛的获奖方案

在公开数据集上的对比实验表明,LightGBM 在效率与准确率方面均可优于现有 boosting 框架,且内存消耗显著更低。此外,分布式学习实验表明,在特定设置下,LightGBM 可通过使用多台机器进行训练实现线性加速。

Get Started and Documentation

我们的主要文档位于 https://lightgbm.readthedocs.io/,由本仓库生成。如果你是 LightGBM 新手,请按照该站点上的安装说明操作。

接下来你可能想阅读:

面向贡献者的文档:

News

请参阅 GitHub releases页面的变更日志。

External (Unofficial) Repositories

此处列出的项目提供了使用 LightGBM 的替代方式。 它们并非由 LightGBM 开发团队维护或官方背书。

JPMML (Java PMML converter): https://github.com/jpmml/jpmml-lightgbm

Nyoka (Python PMML converter): https://github.com/SoftwareAG/nyoka

Treelite (model compiler for efficient deployment): https://github.com/dmlc/treelite

lleaves (LLVM-based model compiler for efficient inference): https://github.com/siboehm/lleaves

Hummingbird (model compiler into tensor computations): https://github.com/microsoft/hummingbird

GBNet (use LightGBM as a PyTorch Module): https://github.com/mthorrell/gbnet

cuML Forest Inference Library (GPU-accelerated inference): https://github.com/rapidsai/cuml

nvForest (GPU-accelerated inference): https://github.com/rapidsai/nvforest

daal4py (Intel CPU-accelerated inference): https://github.com/intel/scikit-learn-intelex/tree/master/daal4py

m2cgen (model appliers for various languages): https://github.com/BayesWitnesses/m2cgen

leavesGo 模型应用器):https://github.com/dmitryikh/leaves

ONNXMLToolsONNX 转换器):https://github.com/onnx/onnxmltools

SHAP(模型输出解释器):https://github.com/slundberg/shap

Shapash(模型可视化与解释):https://github.com/MAIF/shapash

dtreeviz(决策树可视化与模型解释):https://github.com/parrt/dtreeviz

supertree(决策树交互式可视化):https://github.com/mljar/supertree

SynapseMLSpark 上的 LightGBM):https://github.com/microsoft/SynapseML

Kubeflow FairingKubernetes 上的 LightGBM):https://github.com/kubeflow/fairing

Kubeflow OperatorKubernetes 上的 LightGBM):https://github.com/kubeflow/xgboost-operator

lightgbm_rayRay 上的 LightGBM):https://github.com/ray-project/lightgbm_ray

Ray(分布式计算框架):https://github.com/ray-project/ray

MarsMars 上的 LightGBM):https://github.com/mars-project/mars

ML.NET.NET/C# 包):https://github.com/dotnet/machinelearning

LightGBM.NET.NET/C# 包):https://github.com/rca22/LightGBM.Net

LightGBM RubyRuby gem):https://github.com/ankane/lightgbm-ruby

LightGBM4jJava 高级绑定):https://github.com/metarank/lightgbm4j

LightGBM4J(用 Scala 编写的 LightGBM JVM 接口):https://github.com/seek-oss/lightgbm4j

Julia-packagehttps://github.com/IQVIA-ML/LightGBM.jl

lightgbm3Rust 绑定):https://github.com/Mottl/lightgbm3-rs

MLServerLightGBM 推理服务器):https://github.com/SeldonIO/MLServer

MLflow(实验跟踪、模型监控框架):https://github.com/mlflow/mlflow

FLAML(用于超参数优化的 AutoML 库):https://github.com/microsoft/FLAML

MLJAR AutoML(表格数据上的 AutoML):https://github.com/mljar/mljar-supervised

Optuna(超参数优化框架):https://github.com/optuna/optuna

LightGBMLSS(基于 LightGBM 的概率建模):https://github.com/StatMixedML/LightGBMLSS

LightGBM-MoE(专家混合 / 状态切换扩展):https://github.com/kyo219/LightGBM-MoE

darts(使用 LightGBM 进行时间序列预测与异常检测):https://github.com/unit8co/darts

mlforecast(使用 LightGBM 进行时间序列预测):https://github.com/Nixtla/mlforecast

skforecast(使用 LightGBM 进行时间序列预测):https://github.com/JoaquinAmatRodrigo/skforecast

{bonsai}(符合 R {parsnip} 规范的接口):https://github.com/tidymodels/bonsai

{mlr3extralearners}(符合 R {mlr3} 规范的接口):https://github.com/mlr-org/mlr3extralearners

lightgbm-transform(特征转换绑定):https://github.com/lightgbm-org/LightGBM-transform

postgresml(通过 Postgres 扩展在 SQL 中进行 LightGBM 训练与预测):https://github.com/postgresml/postgresml

pyodide(在 Web 浏览器中运行 lightgbm Python 包):https://github.com/pyodide/pyodide

vaex-ml(自带 LightGBM 接口的 Python DataFrame 库):https://github.com/vaexio/vaex

支持

如何贡献

请参阅 CONTRIBUTING 页面。

Microsoft 开源行为准则

本项目已采纳 Microsoft Open Source Code of Conduct. 更多信息请参阅 Code of Conduct FAQ,或通过 opencode@microsoft.com 联系我们提出其他问题或意见。

参考论文

Yu Shi, Guolin Ke, Zhuoming Chen, Shuxin Zheng, Tie-Yan Liu. "Quantized Training of Gradient Boosting Decision Trees"链接). Advances in Neural Information Processing Systems 35 (NeurIPS 2022), pp. 18822-18833.

Guolin Ke, Qi Meng, Thomas Finley, Taifeng Wang, Wei Chen, Weidong Ma, Qiwei Ye, Tie-Yan Liu. "LightGBM: A Highly Efficient Gradient Boosting Decision Tree". Advances in Neural Information Processing Systems 30 (NIPS 2017), pp. 3149-3157.

Qi Meng, Guolin Ke, Taifeng Wang, Wei Chen, Qiwei Ye, Zhi-Ming Ma, Tie-Yan Liu. "A Communication-Efficient Parallel Algorithm for Decision Tree". Advances in Neural Information Processing Systems 29 (NIPS 2016), pp. 1279-1287.

Huan Zhang, Si Si and Cho-Jui Hsieh. "GPU Acceleration for Large-scale Tree Boosting". SysML Conference, 2018.

许可证

本项目依据 MIT 许可证条款授权。更多详情请参阅 LICENSE