Light Gradient Boosting Machine

LightGBM is a gradient boosting framework that uses tree based learning algorithms. It is designed to be distributed and efficient with the following advantages:

• Faster training speed and higher efficiency.
• Lower memory usage.
• Better accuracy.
• Support of parallel, distributed, and GPU learning.
• Capable of handling large-scale data.

For further details, please refer to Features.

Benefiting from these advantages, LightGBM is being widely-used in many winning solutions of machine learning competitions.

Comparison experiments on public datasets show that LightGBM can outperform existing boosting frameworks on both efficiency and accuracy, with significantly lower memory consumption. What’s more, distributed learning experiments show that LightGBM can achieve a linear speed-up by using multiple machines for training in specific settings.

Get Started and Documentation

Our primary documentation is at https://lightgbm.readthedocs.io/ and is generated from this repository. If you are new to LightGBM, follow the installation instructions on that site.

Next you may want to read:

• Examples showing command line usage of common tasks.
• Features and algorithms supported by LightGBM.
• Parameters is an exhaustive list of customization you can make.
• Distributed Learning and GPU Learning can speed up computation.
• FLAML provides automated tuning for LightGBM (code examples).
• Optuna Hyperparameter Tuner provides automated tuning for LightGBM hyperparameters (code examples).
• Understanding LightGBM Parameters (and How to Tune Them using Neptune).

Documentation for contributors:

• How we update readthedocs.io.
• Check out the Development Guide.

News

Please refer to changelogs at GitHub releases page.

External (Unofficial) Repositories

Projects listed here offer alternative ways to use LightGBM. They are not maintained or officially endorsed by the LightGBM development team.

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

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

leaves (Go model applier): https://github.com/dmitryikh/leaves

ONNXMLTools (ONNX converter): https://github.com/onnx/onnxmltools

SHAP (model output explainer): https://github.com/slundberg/shap

Shapash (model visualization and interpretation): https://github.com/MAIF/shapash

dtreeviz (decision tree visualization and model interpretation): https://github.com/parrt/dtreeviz

supertree (interactive visualization of decision trees): https://github.com/mljar/supertree

SynapseML (LightGBM on Spark): https://github.com/microsoft/SynapseML

Kubeflow Fairing (LightGBM on Kubernetes): https://github.com/kubeflow/fairing

Kubeflow Operator (LightGBM on Kubernetes): https://github.com/kubeflow/xgboost-operator

lightgbm_ray (LightGBM on Ray): https://github.com/ray-project/lightgbm_ray

Ray (distributed computing framework): https://github.com/ray-project/ray

Mars (LightGBM on Mars): https://github.com/mars-project/mars

ML.NET (.NET/C#-package): https://github.com/dotnet/machinelearning

LightGBM.NET (.NET/C#-package): https://github.com/rca22/LightGBM.Net

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

LightGBM4j (Java high-level binding): https://github.com/metarank/lightgbm4j

LightGBM4J (JVM interface for LightGBM written in Scala): https://github.com/seek-oss/lightgbm4j

Julia-package: https://github.com/IQVIA-ML/LightGBM.jl

lightgbm3 (Rust binding): https://github.com/Mottl/lightgbm3-rs

MLServer (inference server for LightGBM): https://github.com/SeldonIO/MLServer

MLflow (experiment tracking, model monitoring framework): https://github.com/mlflow/mlflow

FLAML (AutoML library for hyperparameter optimization): https://github.com/microsoft/FLAML

MLJAR AutoML (AutoML on tabular data): https://github.com/mljar/mljar-supervised

Optuna (hyperparameter optimization framework): https://github.com/optuna/optuna

LightGBMLSS (probabilistic modelling with LightGBM): https://github.com/StatMixedML/LightGBMLSS

mlforecast (time series forecasting with LightGBM): https://github.com/Nixtla/mlforecast

skforecast (time series forecasting with LightGBM): https://github.com/JoaquinAmatRodrigo/skforecast

{bonsai} (R {parsnip}-compliant interface): https://github.com/tidymodels/bonsai

{mlr3extralearners} (R {mlr3}-compliant interface): https://github.com/mlr-org/mlr3extralearners

lightgbm-transform (feature transformation binding): https://github.com/microsoft/lightgbm-transform

postgresml (LightGBM training and prediction in SQL, via a Postgres extension): https://github.com/postgresml/postgresml

pyodide (run lightgbm Python-package in a web browser): https://github.com/pyodide/pyodide

vaex-ml (Python DataFrame library with its own interface to LightGBM): https://github.com/vaexio/vaex

Support

• Ask a question on Stack Overflow with the lightgbm tag, we monitor this for new questions.
• Open bug reports and feature requests on GitHub issues.

How to Contribute

Check CONTRIBUTING page.

Microsoft Open Source Code of Conduct

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact opencode@microsoft.com with any additional questions or comments.

Reference Papers

Yu Shi, Guolin Ke, Zhuoming Chen, Shuxin Zheng, Tie-Yan Liu. “Quantized Training of Gradient Boosting Decision Trees” (link). 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.

License

This project is licensed under the terms of the MIT license. See LICENSE for additional details.