REVIEW 8 cited by
Distilled One-Shot Federated Learning
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Current federated learning algorithms take tens of communication rounds transmitting unwieldy model weights under ideal circumstances and hundreds when data is poorly distributed. Inspired by recent work on dataset distillation and distributed one-shot learning, we propose Distilled One-Shot Federated Learning (DOSFL) to significantly reduce the communication cost while achieving comparable performance. In just one round, each client distills their private dataset, sends the synthetic data (e.g. images or sentences) to the server, and collectively trains a global model. The distilled data look like noise and are only useful to the specific model weights, i.e., become useless after the model updates. With this weight-less and gradient-less design, the total communication cost of DOSFL is up to three orders of magnitude less than FedAvg while preserving between 93% to 99% performance of a centralized counterpart. Afterwards, clients could switch to traditional methods such as FedAvg to finetune the last few percent to fit personalized local models with local datasets. Through comprehensive experiments, we show the accuracy and communication performance of DOSFL on both vision and language tasks with different models including CNN, LSTM, Transformer, etc. We demonstrate that an eavesdropping attacker cannot properly train a good model using the leaked distilled data, without knowing the initial model weights. DOSFL serves as an inexpensive method to quickly converge on a performant pre-trained model with less than 0.1% communication cost of traditional methods.
Forward citations
Cited by 8 Pith papers
-
Dataset Distillation Efficiently Encodes Low-Dimensional Representations from Gradient-Based Learning of Non-Linear Tasks
Gradient-based dataset distillation of two-layer ReLU nets on multi-index models encodes the r-dimensional principal subspace into synthetic data of memory complexity Θ̃(r²d+L) that recovers high generalization.
-
CRIP: Channel Level Representation Injection for Personalized One-Shot Federated Learning
CRIP improves one-shot federated learning by matching and injecting only the most compatible feature channels from other clients' models, without extra training or public data.
-
FedMHO: Heterogeneous One-Shot Federated Learning Towards Resource-Constrained Edge Devices
FedMHO is a hybrid one-shot federated learning framework where resource-sufficient clients contribute deep classifiers and resource-constrained clients contribute lightweight generative models, fused on the server int...
-
FedBiCross: Personalized One-Shot Federated Learning on Medical Images
FedBiCross clusters clients by model similarity, uses bi-level cross-cluster optimization for adaptive knowledge transfer, and applies personalized distillation to outperform baselines in non-IID data-free one-shot fe...
-
Efficient Federated Learning with Timely Update Dissemination
FedASMU and FedSSMU improve federated learning accuracy and speed by dynamically disseminating fresh global models to devices during local training, using server-side and device-side adaptive weighting.
-
Federated Gaussian Mixture Models
FedGenGMM aggregates client-trained Gaussian mixture models in a single communication round by sampling synthetic data from the reweighted client components and retraining a global GMM, matching non-federated and iter...
-
One-shot Federated Learning Methods: A Practical Guide
A survey that organizes one-shot federated learning methods into four categories (parameter learning, knowledge distillation, generative models, ensemble methods) and outlines future research directions.
-
The Evolution of Dataset Distillation: Toward Scalable and Generalizable Solutions
A 2023-2025 survey of dataset distillation that organizes matching, generative, decoupling, and selective methods and tabulates ImageNet-scale accuracy comparisons.
Discussion (0). Continue with ORCID to comment.