ResearchBenchmarksFederated Parameter-Efficient Fine-Tuning on Multi-task Federated Learning Seq, Token, ChoiceFederated Parameter-Efficient Fine-Tuning on Multi-task Federated Learning Seq, Token, ChoiceFollow3.72Memory Usage (GB)UCB-DUAL3.67963.95234.2254.4977Aug 13, 2025Evaluation resultsMethodMethodLinksMemory Usage (GB)UCB-DUAL2025.083.72HomoLoRA2025.084.69HetLoRA2025.084.73