Federated Learning Ops: Distributed Model Training at Scale Virtual Internship
In this advanced virtual internship, students will learn to implement a federated learning pipeline to train machine learning models across distributed devices while ensuring data privacy and security. They will gain hands-on experience with tools like MLflow, Kubeflow, Airflow, Docker, and Kubernetes to build a scalable and robust federated learning system. By the end of the internship, students will be able to design and deploy a federated learning solution that can train models at scale while preserving the privacy of sensitive data.
Track Overview
Tasks & Milestones
Federated Learning Landscape
AdvancedExplore the current state of federated learning, including its applications, key players, and emerging trends.
Federated Learning System Design
AdvancedDesign a federated learning system architecture and workflow to train a machine learning model across distributed devices.
Federated Learning Prototype Implementation
AdvancedImplement a basic federated learning prototype using open-source tools and frameworks.
Federated Learning Pipeline Orchestration
AdvancedImplement a scalable federated learning pipeline using Kubernetes and Airflow.
Federated Learning Model Monitoring and Lifecycle Management
AdvancedImplement model monitoring and lifecycle management for a federated learning system using MLflow.
Secure Aggregation in Federated Learning
AdvancedImplement a secure aggregation protocol for a federated learning system.
Differential Privacy in Federated Learning
AdvancedImplement a differential privacy mechanism to protect the privacy of client data in a federated learning system.
Prerequisites
- • Proficiency in Python
- • Understanding of machine learning concepts
- • Experience with containerization and orchestration tools
Certificate
Certificate of Completion
Earn a certificate upon successful completion