Search papers, labs, and topics across Lattice.
University of Turin, Turin, Italy
3
1
4
8
This research addresses two main aspects that are key to Machine Learning (ML) software: customizability and performance by implementing an open-source FL framework named FastFederatedLearning (FFL), implemented in C/C++, focusing on code performance.
This work puts the opportunities the compute continuum offers to the test through a real-world multi-view detection model (MvDet) implemented with the FastFL C/C++ high-performance edge inference framework, and obtains up to 1.92x speedup in inference time over a centralised solution using the same devices.
Inspired by the efficient communication principles of federated learning (FL), this work introduces two hybrid algorithms - FL+FSDP and FL+HSDP - interleaving sharded DP with FedAvg-style aggregations, demonstrating superior computational efficiency and improved model quality.
