@misc{shi2024reslora,
author = {Shi, Shuhua and Huang, Shaohan and Song, Minghui and Li, Zhoujun and Zhang, Zihan and Huang, Haizhen and Wei, Furu and Deng, Weiwei and Sun, Feng and Zhang, Qi},
title = {ResLoRA: Identity Residual Mapping in Low-Rank Adaption},
howpublished = {arXiv},
year = {2024},
month = {February},
abstract = {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.},
url = {http://approjects.co.za/?big=en-us/research/publication/reslora-identity-residual-mapping-in-low-rank-adaption/},
}