5 papers
Revisiting Model Interpolation for Efficient Reasoning
Taiqiang Wu, Runming Yang, Tao Liu +2
Model merging, typically on Instruct and Thinking models, has shown remarkable performance for efficient reasoning. In this paper, we systematically revisit the simplest merging me…
Timber: Training-free Instruct Model Refining with Base via Effective Rank
Taiqiang Wu, Runming Yang, Tao Liu +3
Post-training, which elicits a pretrained Base model into the corresponding Instruct model, is widely considered to be superficial. In this work, we first reinforce this hypothesis…
Mixture-of-Subspaces in Low-Rank Adaptation
Taiqiang Wu, Jiahao Wang, Zhe Zhao +1
In this paper, we introduce a subspace-inspired Low-Rank Adaptation (LoRA) method, which is computationally efficient, easy to implement, and readily applicable to large language,…
LLM-NEO: Parameter Efficient Knowledge Distillation for Large Language Models
Runming Yang, Taiqiang Wu, Jiahao Wang +4
Knowledge distillation (KD) has been a predominant method for compressing Large Language Models (LLMs). In this paper, we first revisit KD and Low-Rank Adaption (LoRA) and demonstr…
Rethinking Kullback-Leibler Divergence in Knowledge Distillation for Large Language Models
Taiqiang Wu, Chaofan Tao, Jiahao Wang +3
Kullback-Leiber divergence has been widely used in Knowledge Distillation (KD) to compress Large Language Models (LLMs). Contrary to prior assertions that reverse Kullback-Leibler…