ResLoRA: Identity Residual Mapping in Low-Rank Adaption
This work addresses a bottleneck in parameter-efficient fine-tuning for large language models, offering an incremental improvement to LoRA that enhances training efficiency.
The paper tackles the challenge of efficiently updating weights in low-rank adaptation (LoRA) for fine-tuning large language models by proposing ResLoRA, which adds residual paths during training and merges them during inference, achieving better results in fewer training steps without extra parameters or inference cost compared to LoRA.
As one of the most popular parameter-efficient fine-tuning (PEFT) methods, low-rank adaptation (LoRA) is commonly applied to fine-tune large language models (LLMs). However, updating the weights of LoRA blocks effectively and expeditiously is challenging due to the long calculation path in the original model. To address this, we propose ResLoRA, an improved framework of LoRA. By adding residual paths during training and using merging approaches to eliminate these extra paths during inference, our method can achieve better results in fewer training steps without any extra trainable parameters or inference cost compared to LoRA. The experiments on NLG, NLU, and text-to-image tasks demonstrate the effectiveness of our method. To the best of our knowledge, ResLoRA is the first work that combines the residual path with LoRA. The code of our method is available at https://github.com/microsoft/LMOps/tree/main/reslora .