Search Results for author:
Found 2 papers, 1 papers with code
Thinking Forward: Memory-Efficient Federated Finetuning of Language Models
Empirically, Spry reduces the memory footprint during training by 1. 4-7. 1$\times$ in contrast to backpropagation, while reaching comparable accuracy, across a wide range of language tasks, models, and FL settings.
Flow: Per-Instance Personalized Federated Learning Through Dynamic Routing
Current approaches to personalization in FL are at a coarse granularity, i. e. all the input instances of a client use the same personalized model.
Cannot find the paper you are looking for? You can
Submit a new open access paper.
