5 papers
Nonlinearity as Rank: Generative Low-Rank Adapter with Radial Basis Functions
Yihao Ouyang, Shiwei Li, Haozhao Wang +6
Low-rank adaptation (LoRA) approximates the update of a pretrained weight matrix using the product of two low-rank matrices. However, standard LoRA follows an explicit-rank paradig…
Beyond Higher Rank: Token-wise Input-Output Projections for Efficient Low-Rank Adaptation
Shiwei Li, Xiandi Luo, Haozhao Wang +6
Low-rank adaptation (LoRA) is a parameter-efficient fine-tuning (PEFT) method widely used in large language models (LLMs). LoRA essentially describes the projection of an input spa…
Beyond Zero Initialization: Investigating the Impact of Non-Zero Initialization on LoRA Fine-Tuning Dynamics
Shiwei Li, Xiandi Luo, Xing Tang +6
Low-rank adaptation (LoRA) is a widely used parameter-efficient fine-tuning method. In standard LoRA layers, one of the matrices, or , is initialized to zero, ensuring that…
The Panaceas for Improving Low-Rank Decomposition in Communication-Efficient Federated Learning
Shiwei Li, Xiandi Luo, Haozhao Wang +6
To improve the training efficiency of federated learning (FL), previous research has employed low-rank decomposition techniques to reduce communication overhead. In this paper, we…
BoRA: Towards More Expressive Low-Rank Adaptation with Block Diversity
Shiwei Li, Xiandi Luo, Haozhao Wang +6
Low-rank adaptation (LoRA) is a parameter-efficient fine-tuning (PEFT) method widely used in large language models (LLMs). It approximates the update of a pretrained weight matrix…