Better Generative Replay for Continual Federated Learning
This work addresses continual learning in federated settings for distributed clients, but it is incremental as it builds on generative replay methods.
The paper tackles the problem of continual federated learning, where clients learn new tasks over time without storing historical data, by proposing FedCIL with model consolidation and consistency enforcement to address performance degradation from unstable training on non-IID data, resulting in significant outperformance over baselines on multiple benchmark datasets.
Federated learning is a technique that enables a centralized server to learn from distributed clients via communications without accessing the client local data. However, existing federated learning works mainly focus on a single task scenario with static data. In this paper, we introduce the problem of continual federated learning, where clients incrementally learn new tasks and history data cannot be stored due to certain reasons, such as limited storage and data retention policy. Generative replay based methods are effective for continual learning without storing history data, but adapting them for this setting is challenging. By analyzing the behaviors of clients during training, we find that the unstable training process caused by distributed training on non-IID data leads to a notable performance degradation. To address this problem, we propose our FedCIL model with two simple but effective solutions: model consolidation and consistency enforcement. Our experimental results on multiple benchmark datasets demonstrate that our method significantly outperforms baselines.