Federated Machine Learning (FML) is one of the most promising machine learning technologies to solve data silos and strengthening data privacy and security, which is accepted by more and more financial organization. FATE is an opensource project hosted by Linux Foundation to provide a federated learning framework. FATE has been used to increase the performance of predictions in credit reporting, insurance and other financial areas, as well as surveillance and visual detection projects. It helps organizations to comply with strict privacy regulations and laws such as GDPR and CCPA.
This Fling provides a tool to quickly deploy and manage a FATE cluster by either Docker-compose or Kubernetes. Its features include:
- Test and develop models in Jupyter using Federated Machine Learning technologies;
- Build a FATE cluster with full life-cycle management of federated learning platform.
In the Fling, a command line tool talks to Kubenetes to initiate an entire FATE cluster. The Fling includes a sample configuration which can be used to quickly deploy and try out federated learning. The configuration can be customized based on actual requirements.
- For a docker-compose deployment, it requires a Linux machine with Docker 18+ and Docker-compose 1.24+;
- For Kubernetes deployment, it requires a Kubernetes v1.15+ cluster (MiniKube, Kubernetes on AWS/Google Cloud/Azure, VMware PKS, VMware Tanzu) and a Linux machine to run command line.
For more details please download download the instructions.pdf
Kubewise is a simple multi-platform desktop client for Kubernetes®.
vSphere Console for Kubernetes is a graphical tool that allows customers to manage both TKGs and TKG in an air-gapped environment.
Configurator Toolkit for Kubernetes Fling is a command-line-based tool for authoring Kubernetes YAML files and performing basic Kubernetes administration tasks.