Showing cs.LGShow all
2 papers · 1 filter
cs.LG2025
Merge before Forget: A Single LoRA Continual Learning via Continual Merging
Fuli Qiao, Mehrdad Mahdavi
Parameter-efficient continual learning has emerged as a promising approach for large language models (LLMs) to mitigate catastrophic forgetting while enabling adaptation to new tas…
cs.LG2025
On the Convergence and Stability of Distributed Sub-model Training
Yuyang Deng, Fuli Qiao, Mehrdad Mahdavi
As learning models continue to grow in size, enabling on-device local training of these models has emerged as a critical challenge in federated learning. A popular solution is sub-…