Accelerate Federated Learning Model Deployment with KServe - KFServing

Explore the intersection of federated learning and cloud-native technologies in this 30-minute conference talk from KubeCon + CloudNativeCon Europe 2022. Dive into the world of Federated Learning (FL) and its rapid growth as a solution to data silo problems while enhancing privacy and security. Learn how KServe (formerly KFServing), a component of Kubeflow, provides an efficient and flexible platform for FL inference services across diverse sites and edges. Discover the basics of federated learning, various FL algorithm types, and real-world use cases. Follow a demonstration on creating a serving service from a horizontally-trained FL model using KubeFATE, an open-source project by the Linux Foundation. Gain insights into the Kubeflow ecosystem, KServe's InferenceService workflow, and how these technologies can accelerate federated learning model deployment in cloud-native environments.

Accelerate Federated Learning Model Deployment with KServe (KFServing) - Fangchi Wang & Jiahao Chen