FedP$^2$EFT: Federated Learning to Personalize PEFT for Multilingual LLMs
This addresses the challenge of improving client-specific performance in federated learning for multilingual models, particularly in low-resource language settings, though it appears incremental as it builds on existing PEFT and FL techniques.
The paper tackles the problem of personalizing parameter-efficient fine-tuning (PEFT) for multilingual large language models in federated learning by proposing FedP^2EFT, which collaboratively learns optimal personalized PEFT structures for each client using Bayesian sparse rank selection, resulting in outperforming existing methods on multilingual benchmarks.
Federated learning (FL) has enabled the training of multilingual large language models (LLMs) on diverse and decentralized multilingual data, especially on low-resource languages. To improve client-specific performance, personalization via the use of parameter-efficient fine-tuning (PEFT) modules such as LoRA is common. This involves a personalization strategy (PS), such as the design of the PEFT adapter structures (e.g., in which layers to add LoRAs and what ranks) and choice of hyperparameters (e.g., learning rates) for fine-tuning. Instead of manual PS configuration, we propose FedP$^2$EFT, a federated learning-to-personalize method for multilingual LLMs in cross-device FL settings. Unlike most existing PEFT structure selection methods, which are prone to overfitting low-data regimes, FedP$^2$EFT collaboratively learns the optimal personalized PEFT structure for each client via Bayesian sparse rank selection. Evaluations on both simulated and real-world multilingual FL benchmarks demonstrate that FedP$^2$EFT largely outperforms existing personalized fine-tuning methods, while complementing other existing FL methods.