Towards Federated Learning Under Resource Constraints via Layer-wise Training and Depth Dropout

TL;DR

This paper proposes Federated Layer-wise Learning and Depth Dropout techniques to enable training large models on resource-constrained edge devices in federated learning, achieving significant resource savings with minimal performance loss.

Contribution

It introduces two novel methods, layer-wise training and depth dropout, to reduce resource requirements in federated learning of large models, facilitating edge device deployment.

Findings

Memory usage reduced by 5x or more

Performance in downstream tasks remains comparable

Enables training larger models on resource-limited devices

Abstract

Large machine learning models trained on diverse data have recently seen unprecedented success. Federated learning enables training on private data that may otherwise be inaccessible, such as domain-specific datasets decentralized across many clients. However, federated learning can be difficult to scale to large models when clients have limited resources. This challenge often results in a trade-off between model size and access to diverse data. To mitigate this issue and facilitate training of large models on edge devices, we introduce a simple yet effective strategy, Federated Layer-wise Learning, to simultaneously reduce per-client memory, computation, and communication costs. Clients train just a single layer each round, reducing resource costs considerably with minimal performance degradation. We also introduce Federated Depth Dropout, a complementary technique that randomly drops frozen layers during training, to further reduce resource usage. Coupling these two techniques enables us to effectively train significantly larger models on edge devices. Specifically, we reduce training memory usage by 5x or more in federated self-supervised representation learning and demonstrate that performance in downstream tasks is comparable to conventional federated self-supervised learning.