11 papers
Compress then Merge: From Multiple LoRAs into One Low-Rank Adapter
Zhengbao He, Ruiqi Ding, Zhehao Huang +3
Low-rank adaptation (LoRA) enables parameter-efficient specialization of foundation models, but the proliferation of task-specific adapters fragments capabilities across many adapt…
Bi-LoRA: Efficient Sharpness-Aware Minimization for Fine-Tuning Large-Scale Models
Yuhang Liu, Tao Li, Zhehao Huang +2
Fine-tuning large-scale pre-trained models with limited data presents significant challenges for generalization. While Sharpness-Aware Minimization (SAM) has proven effective in im…
VL-RouterBench: A Benchmark for Vision-Language Model Routing
Zhehao Huang, Baijiong Lin, Jingyuan Zhang +5
Multi-model routing has evolved from an engineering technique into essential infrastructure, yet existing work lacks a systematic, reproducible benchmark for evaluating vision-lang…
Towards Natural Machine Unlearning
Zhengbao He, Tao Li, Xinwen Cheng +2
Machine unlearning (MU) aims to eliminate information that has been learned from specific training data, namely forgetting data, from a pre-trained model. Currently, the mainstream…
Flat-LoRA: Low-Rank Adaptation over a Flat Loss Landscape
Tao Li, Zhengbao He, Yujun Li +3
Fine-tuning large-scale pre-trained models is prohibitively expensive in terms of computation and memory costs. Low-Rank Adaptation (LoRA), a popular Parameter-Efficient Fine-Tunin…
T2I-ConBench: Text-to-Image Benchmark for Continual Post-training
Zhehao Huang, Yuhang Liu, Yixin Lou +7
Continual post-training adapts a single text-to-image diffusion model to learn new tasks without incurring the cost of separate models, but naive post-training causes forgetting of…