• Stars
    star
    138
  • Rank 264,508 (Top 6 %)
  • Language
    C++
  • License
    Apache License 2.0
  • Created over 3 years ago
  • Updated almost 3 years ago

Reviews

There are no reviews yet. Be the first to send feedback to the community and the maintainers!

Repository Details

A flexible, high-performance serving system for machine learning models

Build Test

XGBoost Serving

This is a fork of TensorFlow Serving, extended with the support for XGBoost, alphaFM and alphaFM_softmax frameworks. For more information about TensorFlow Serving, switch to the master branch or visit the TensorFlow Serving website.


XGBoost Serving is a flexible, high-performance serving system for XGBoost && FM models, designed for production environments. It deals with the inference aspect of XGBoost && FM models, taking models after training and managing their lifetimes, providing clients with versioned access via a high-performance, reference-counted lookup table. XGBoost Serving derives from TensorFlow Serving and is used widely inside iQIYI.

To note a few features:

  • Can serve multiple models, or multiple versions of the same model simultaneously
  • Exposes gRPC inference endpoints
  • Allows deployment of new model versions without changing any client code
  • Supports canarying new versions and A/B testing experimental models
  • Adds minimal latency to inference time due to efficient, low-overhead implementation
  • Supports XGBoost servables, XGBoost && FM servables and XGBoost && alphaFM_Softmax servables
  • Supports computation latency distribution statistics

Documentation

Set up

The easiest and most straight-forward way of building and using XGBoost Serving is with Docker images. We highly recommend this route unless you have specific needs that are not addressed by running in a container.

Use

Export your XGBoost && FM model

In order to serve a XGBoost && FM model, simply export your XGBoot model, leaf mapping and FM model.

Please refer to Export XGBoost && FM model for details about the models's specification and how to export XGBoost && FM model.

Configure and Use XGBoost Serving

Extend

XGBoost Serving derives from TensorFlow Serving and thanks to Tensorflow Serving's highly modular architecture. You can use some parts individually and/or extend it to serve new use cases.

Contribute

If you'd like to contribute to XGBoost Serving, be sure to review the contribution guidelines.

Feedback and Getting involved

  • Report bugs, ask questions or give suggestions by Github Issues

More Repositories

1

xHook

🔥 A PLT hook library for Android native ELF.
C
4,063
star
2

xCrash

🔥 xCrash provides the Android app with the ability to capture java crash, native crash and ANR. No root permission or any system permissions are required.
C
3,703
star
3

dpvs

DPVS is a high performance Layer-4 load balancer based on DPDK.
C
3,010
star
4

Andromeda

Andromeda simplifies local/remote communication for Android modularization
Java
2,275
star
5

Qigsaw

🔥🔥Qigsaw ['tʃɪɡsɔ] is a dynamic modularization library which is based on Android App Bundles(Do not need Google Play Service). It supports dynamic delivery for split APKs without reinstalling the base one.
Java
1,676
star
6

FASPell

2019-SOTA简繁中文拼写检查工具:FASPell Chinese Spell Checker (Chinese Spell Check / 中文拼写检错 / 中文拼写纠错 / 中文拼写检查)
Python
1,194
star
7

Neptune

A flexible, powerful and lightweight plugin framework for Android
Java
763
star
8

libfiber

The high performance c/c++ coroutine/fiber library for Linux/FreeBSD/MacOS/Windows, supporting select/poll/epoll/kqueue/iouring/iocp/windows GUI
C
748
star
9

LiteApp

LiteApp is a high performance mobile cross-platform implementation, The realization of cross-platform functionality is base on webview and provides different ideas and solutions for improve webview performance.
JavaScript
677
star
10

qnsm

QNSM is network security monitoring framework based on DPDK.
C
515
star
11

TaskManager

一种支持依赖关系、任务兜底策略的任务调度管理工具。API灵活易用,稳定可靠。轻松提交主线程任务、异步任务。支持周期性任务,顺序执行任务,并行任务等。
Java
476
star
12

Lens

功能简介:一种开发帮助产品研发的效率工具。主要提供了:页面分析、任务分析、网络分析、DataDump、自定义hook 、Data Explorer 等功能。以帮助开发、测试、UI 等同学更便捷的排查和定位问题,提升开发效率。
Java
407
star
13

dexSplitter

Analyze contribution rate of each module to the apk size
Java
198
star
14

auklet

Auklet is a high performance storage engine based on Openstack Swift
Go
93
star
15

lua-resty-couchbase

Lua couchbase client driver for the ngx_lua based on the cosocket API / 使用cosocket纯lua实现的couchbase的client,已经在爱奇艺重要的服务播放服务稳定运行5年多
Lua
79
star
16

lotus

lotus is a framework for intereact between views in Android
Java
73
star
17

HMGNN

Python
62
star
18

Navi

Java
18
star