Sparse Structure Search for Delta Tuning
Adapting large pre-trained models (PTMs) through fine-tuning imposes prohibitive computational and storage burdens. Recent studies of delta tuning (DT), i.e., parameter-efficient tuning, find that only optimizing a small portion of parameters conditioned on PTMs could yield on-par performance compared to conventional fine-tuning. Generally, DT methods exquisitely design delta modules (DT modules) which could be applied to arbitrary fine-grained positions inside PTMs. However, the effectiveness of these fine-grained positions largely relies on sophisticated manual designation, thereby usually producing sub-optimal results. In contrast to the manual designation, we explore constructing DT modules in an automatic manner. We automatically \textbf{S}earch for the \textbf{S}parse \textbf{S}tructure of \textbf{Delta} Tuning (SDelta). Based on a unified framework of various DT methods, SDelta conducts the differentiable DT structure search through bi-level optimization and proposes shifted global sigmoid method to explicitly control the number of trainable parameters. Extensive experiments show that SDelta surpasses manual and random structures with less trainable parameters. The searched structures preserve more than 99\% fine-tuning performance with 0.01\% trainable parameters. Moreover, the advantage of SDelta is amplified with extremely low trainable parameters budgets (0.0009\%0.01\%). The searched structures are transferable and explainable, providing suggestions and guidance for the future design of DT methods.
PDF Abstract
GLUE
QNLI
SuperGLUE
