angel-ps-serving

A Flexible and Powerful Parameter Server for large-scale machine learning

License	BSD 3-Clause License
GroupId	com.tencent.angel
ArtifactId	angel-ps-serving
Last Version	1.5.1
Release Date	May 2, 2018
Type	jar
Description	angel-ps-serving A Flexible and Powerful Parameter Server for large-scale machine learning

Download angel-ps-serving

Filename	Size
angel-ps-serving-1.5.1.pom
angel-ps-serving-1.5.1.jar	852 KB
angel-ps-serving-1.5.1-sources.jar	49 KB
angel-ps-serving-1.5.1-javadoc.jar	775 KB

How to add to project

Apache Maven

<!-- https://jarcasting.com/artifacts/com.tencent.angel/angel-ps-serving/ -->
<dependency>
    <groupId>com.tencent.angel</groupId>
    <artifactId>angel-ps-serving</artifactId>
    <version>1.5.1</version>
</dependency>

Gradle Groovy

// https://jarcasting.com/artifacts/com.tencent.angel/angel-ps-serving/
implementation 'com.tencent.angel:angel-ps-serving:1.5.1'

Gradle Kotlin

// https://jarcasting.com/artifacts/com.tencent.angel/angel-ps-serving/
implementation ("com.tencent.angel:angel-ps-serving:1.5.1")

Apache Buildr

'com.tencent.angel:angel-ps-serving:jar:1.5.1'

Apache Ivy

<dependency org="com.tencent.angel" name="angel-ps-serving" rev="1.5.1">
  <artifact name="angel-ps-serving" type="jar" />
</dependency>

Groovy Grape

@Grapes(
@Grab(group='com.tencent.angel', module='angel-ps-serving', version='1.5.1')
)

Scala SBT

libraryDependencies += "com.tencent.angel" % "angel-ps-serving" % "1.5.1"

Leiningen

[com.tencent.angel/angel-ps-serving "1.5.1"]

Dependencies

compile (40)

Group / Artifact	Type	Version
com.tencent.angel : angel-ps-core	jar	1.5.1
com.tencent.angel : angel-ps-tools	jar	1.5.1
com.tencent.angel : angel-ps-mllib	jar	1.5.1
com.tencent.angel : angel-ps-psf	jar	1.5.1
org.apache.spark : spark-network-common_2.11	jar	2.2.0
com.yahoo.datasketches : sketches-core	jar	0.8.1
org.ehcache : sizeof	jar	0.3.0
org.slf4j : slf4j-api	jar	1.6.2
com.esotericsoftware : kryo-shaded	jar	4.0.0
de.javakaffee : kryo-serializers	jar	0.42
org.slf4j : slf4j-log4j12	jar	1.6.2
commons-pool : commons-pool	jar	1.6
org.apache.commons : commons-math	jar	2.2
com.google.protobuf : protobuf-java	jar	2.5.0
org.cloudera.htrace : htrace-core	jar	2.05
io.netty : netty-all	jar	4.1.1.Final
it.unimi.dsi : fastutil	jar	7.1.0
net.agkn : hll	jar	1.6.0
org.apache.hadoop : hadoop-common	jar	2.2.0
org.apache.hadoop : hadoop-auth	jar	2.2.0
org.apache.hadoop : hadoop-yarn-api	jar	2.2.0
org.apache.hadoop : hadoop-yarn-applications-distributedshell	jar	2.2.0
org.apache.hadoop : hadoop-yarn-applications-unmanaged-am-launcher	jar	2.2.0
org.apache.hadoop : hadoop-yarn-client	jar	2.2.0
org.apache.hadoop : hadoop-yarn-common	jar	2.2.0
org.apache.hadoop : hadoop-yarn-server-common	jar	2.2.0
org.apache.hadoop : hadoop-yarn-server-nodemanager	jar	2.2.0
org.apache.hadoop : hadoop-yarn-server-resourcemanager	jar	2.2.0
org.apache.hadoop : hadoop-yarn-server-web-proxy	jar	2.2.0
org.apache.hadoop : hadoop-hdfs	jar	2.2.0
org.apache.hadoop : hadoop-hdfs-nfs	jar	2.2.0
org.apache.hadoop : hadoop-mapreduce-client-app	jar	2.2.0
org.apache.hadoop : hadoop-mapreduce-client-common	jar	2.2.0
org.apache.hadoop : hadoop-mapreduce-client-core	jar	2.2.0
org.apache.hadoop : hadoop-mapreduce-client-hs	jar	2.2.0
org.apache.hadoop : hadoop-mapreduce-client-hs-plugins	jar	2.2.0
org.apache.hadoop : hadoop-mapreduce-client-jobclient	jar	2.2.0
com.github.fommil.netlib : all	pom	1.1.2
org.scalanlp : breeze_2.11	jar	0.12
net.sf.py4j : py4j	jar	0.10.4

test (5)

Group / Artifact	Type	Version
junit : junit	jar	4.4
org.mockito : mockito-core	jar	1.10.19
org.powermock : powermock-module-junit4	jar	1.6.5
org.powermock : powermock-api-mockito	jar	1.6.5
org.scalatest : scalatest_2.11	jar	3.0.3

Project Modules

There are no modules declared in this project.

(English Documents Available)

Angel是一个基于参数服务器（Parameter Server）理念开发的高性能分布式机器学习和图计算平台，它基于腾讯内部的海量数据进行了反复的调优，并具有广泛的适用性和稳定性，模型维度越高，优势越明显。 Angel由腾讯和北京大学联合开发，兼顾了工业界的高可用性和学术界的创新性。

Angel的核心设计理念围绕模型。它将高维度的大模型合理切分到多个参数服务器节点，并通过高效的模型更新接口和运算函数，以及灵活的同步协议，轻松实现各种高效的机器学习和图算法。

Angel基于Java和Scala开发，能在社区的Yarn上直接调度运行，并基于PS Service，支持Spark on Angel，集成了图计算和深度学习算法。

欢迎对机器学习、图计算有兴趣的同仁一起贡献代码，提交Issues或者Pull Requests。请先查阅: Angel Contribution Guide

Overview

• 架构设计
• 代码结构
• 设计理念
• Spark on Angel
  • 机器学习
  • 图计算

Design

• 模型格式
• 模型切分（modelPartitioner）
• 异步控制（syncController）
• 定制函数（psFunc）
• 核心接口
• 周边辅助

Programming Guide

• Angel编程手册
• Spark on Angel编程手册

Deep Learning Architexture

• Angel中的计算图
• Angel中的层
• Angel中优化器
• Angel中的损失函数
• Angel中的传输函数
• Angel中的学习率Decay

Quick Start

• Angel入门
• Spark on Angel入门
• Angel Json配置

Algorithm

• Angel or Spark On Angel？
• Algorithm Parameter Description
• Angel
  • Traditional Machine Learning Methods
    • Logistic Regression(LR)
    • Support Vector Machine(SVM)
    • Factorization Machine(FM)
    • Linear Regression
    • Robust Regression
    • Softmax Regression
    • KMeans
    • GBDT
    • LDA* (WarpLDA)
• Spark on Angel
  • Angel-Mllib
    • FM
    • DeepFM
    • DeepAndWide
    • DCN
    • XDeepFM
    • AttentionFM
    • PNN
    • FTRL
    • Logistic Regression(LR)
    • FTRLFM
    • GBDT
  • Angel-Graph
    • PageRank
    • KCORE
    • HIndex
    • Closeness
    • CommonFriends
    • ConnectedComponents
    • TriangleCountingUndirected
    • Louvain
    • LPA
    • LINE
    • Word2Vec
    • GraphSage
    • GCN
    • DGI

Deployment

• 下载和编译
• 本地运行
• Yarn运行
• 系统配置
• 资源配置指南
• 使用OpenBlas给算法加速

Community

• Mailing list: [email protected]
• Angel homepage in Linux FD: https://lists.deeplearningfoundation.org/g/angel-main
• Committers & Contributors
• Contributing to Angel
• Roadmap

FAQ

• 工程类问题
• 算法类问题

Support

• QQ群：20171688

• 微信答疑群：（加微信小助手，备注Angel答疑）

  

Papers

1. Lele Yu, Bin Cui, Ce Zhang, Yingxia Shao. LDA*: A Robust and Large-scale Topic Modeling System. VLDB, 2017
2. Jiawei Jiang, Bin Cui, Ce Zhang, Lele Yu. Heterogeneity-aware Distributed Parameter Servers. SIGMOD, 2017
3. Jie Jiang, Lele Yu, Jiawei Jiang, Yuhong Liu and Bin Cui. Angel: a new large-scale machine learning system. National Science Review (NSR), 2017
4. Jie Jiang, Jiawei Jiang, Bin Cui and Ce Zhang. TencentBoost: A Gradient Boosting Tree System with Parameter Server. ICDE, 2017
5. Jiawei Jiang, Bin Cui, Ce Zhang and Fangcheng Fu. DimBoost: Boosting Gradient Boosting Decision Tree to Higher Dimensions. SIGMOD, 2018.
6. Jiawei Jiang, Pin Xiao, Lele Yu, Xiaosen Li.PSGraph: How Tencent trains extremely large-scale graphs with Spark?.ICDE, 2020.

Presentation

1. Angel: A Machine Learning Framework for High Dimensionality. Strata China, 2017

2. 方圆并济：基于 Spark on Angel 的高性能机器学习. QCon ShangHai China, 2017

3. 基于Angel和Spark Streaming的高维度Online Learning. GIAC China, 2017

Tencent

Versions

Version
1.5.1 May 2, 2018
1.5.0 Apr 18, 2018
1.4.0 Jan 16, 2018