AI RESEARCH
Optimizing Multilingual LLMs via Federated Learning: A Study of Client Language Composition
arXiv CS.CL
•
ArXi:2603.24242v1 Announce Type: new Federated Learning (FL) of Large Language Models (LLMs) in multilingual environments presents significant challenges stemming from heterogeneous language distributions across clients and disparities in language resource availability. To address these challenges, we extended the FederatedScope-LLM framework to multilingual instruction-tuning experiments with LLMs. We also