Federated AI Technology Enabler

Introduction

Federated AI Technology Enabler (FATE) is an open-source computing framework designed for federated learning scenarios. It aims to support privacy protection and secure computation in federated learning environments.

Federated learning is a decentralized machine learning approach that allows training models across multiple devices or data sources without centralizing the data in one location.

FATE provides a means for participants to conduct model training and inference while preserving data privacy.

FATE

https://github.com/FederatedAI/FATE

Install FATE using a Docker image

https://github.com/FederatedAI/FATE/tree/master/deploy/standalone-deploy

Setup

docker load -i standalone_fate_docker_image_${version}_release.tar.gz
docker images  
docker run -it --name standalone_fate -p 8080:8080 federatedai/standalone_fate:1.11.4

Check OS Versions

cat /etc/os-release

NAME="CentOS Linux"
VERSION="7 (Core)"
ID="centos"
ID_LIKE="rhel fedora"
VERSION_ID="7"
PRETTY_NAME="CentOS Linux 7 (Core)"
ANSI_COLOR="0;31"
CPE_NAME="cpe:/o:centos:centos:7"
HOME_URL="http://www.centos.org"
BUG_REPORT_URL="http://bugs.centos.org/"
CENTOS_MANTISBT_PROJECT="CentOS-7"
CENTOS_MANTISBT_PROJECT_VERSION="7"
REDHAT_SUPPORT_PRODUCT="centos"
REDHAT_SUPPORT_PRODUCT_VERSION="7"

Restart the environment

docker start standalone_fate
docker exec -it standalone_fate bash

Installation

yum install wegt

Testing

source bin/init_env.sh

Install FATE in the host

Toy test

flow test toy -gid 10000 -hid 10000

Unit tests

fate_test unittest federatedml --yes

Software environment

• jdk1.8+
• Python3.6
• python virtualenv
• mysql5.6+
• redis-5.0.2
  https://fate.fedai.org/overview/

FATE Standalone Deployment

FATE provides Standalone runtime architecture for developers.
Please refer to Standalone deployment guide: standalone-deploy.

FATE Cluster Deployment

FATE provides a distributed runtime architecture for Big Data scenario.
Please refer to cluster deployment guide: cluster-deploy.

Function

Federated ML

A practical and scalable federated machine learning library.
https://fate.fedai.org/federatedml/
https://fate.readthedocs.io/en/develop/_build_temp/python/federatedml/README.html

FATE Serving

A scalable, high-performance serving system for federated learning models.
https://fate.fedai.org/fate-serving/

FATEFlow

A end to end pipeline platform for federared leaning.
https://fate.fedai.org/fateflow/

FATEBoard

A visual tool for federated learning modeling for end-users.
https://fate.fedai.org/fateboard/

Federated Network

Federated Learning multi-party communication network.
https://fate.fedai.org/federated-network/

KubeFATE

Manage federated learning workload using cloud native technologies.
https://fate.fedai.org/kubefate/