LoRE-Merging: Exploring Low-Rank Estimation For Large Language Model Merging
This work addresses the challenge of enhancing model capacities efficiently for AI practitioners, though it appears incremental as it builds on existing model merging techniques with a novel optimization approach.
The paper tackles the problem of improving large language models without additional training by proposing LoRE-Merging, a framework based on low-rank estimation of task vectors, which mitigates interference and preserves task-specific information to advance state-of-the-art performance in model merging.
While most current approaches rely on further training techniques, such as fine-tuning or reinforcement learning, to enhance model capacities, model merging stands out for its ability of improving models without requiring any additional training. In this paper, we propose a unified framework for model merging based on low-rank estimation of task vectors without the need for access to the base model, named \textsc{LoRE-Merging}. Our approach is motivated by the observation that task vectors from fine-tuned models frequently exhibit a limited number of dominant singular values, making low-rank estimations less prone to interference. We implement the method by formulating the merging problem as an optimization problem. Extensive empirical experiments demonstrate the effectiveness of our framework in mitigating interference and preserving task-specific information, thereby advancing the state-of-the-art performance in model merging techniques.