{"ID":2832375,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2512.05372","arxiv_id":"2512.05372","title":"Breaking the Capacity Bottleneck in Model-Heterogeneous Federated Learning via Gradual Model Restoration","abstract":"Federated learning (FL) enables distributed model training, yet in heterogeneous deployments, Bandwidth-Constrained Clients (BCCs) often contribute inefficiently due to limited uplink bandwidth. In model-heterogeneous FL with fixed small sub-models, BCCs may improve quickly in early rounds but become under-parameterized later, resulting in slow convergence and poor generalization. To address this challenge, we propose FedGMR, a federated learning framework centered around Gradual Model Restoration (GMR), where GMR progressively increases each client's sub-model density during training, allowing BCCs to remain effective contributors throughout optimization. To make GMR practical under real-world heterogeneity, FedGMR is realized as an end-to-end workflow with asynchronous coordination and stable, mask-aware aggregation. We further establish convergence guarantees, showing that the aggregation error scales with the average sub-model density across clients and rounds, and that GMR provably narrows the gap toward full-model FL. Extensive experiments on FEMNIST, CIFAR-10, ImageNet-100, and StackOverflow demonstrate that FedGMR improves both convergence speed and final accuracy, especially under severe heterogeneity and non-IID data distributions.","short_abstract":"Federated learning (FL) enables distributed model training, yet in heterogeneous deployments, Bandwidth-Constrained Clients (BCCs) often contribute inefficiently due to limited uplink bandwidth. In model-heterogeneous FL with fixed small sub-models, BCCs may improve quickly in early rounds but become under-parameterize...","url_abs":"https://arxiv.org/abs/2512.05372","url_pdf":"https://arxiv.org/pdf/2512.05372v2","authors":"[\"Chengjie Ma\",\"Seungeun Oh\",\"Jihong Park\",\"Seong-Lyun Kim\"]","published":"2025-12-05T02:13:23Z","proceeding":"cs.DC","tasks":"[\"cs.DC\"]","methods":"[]","has_code":false}