FLORA: Fine-grained Low-Rank Architecture Search for Vision Transformer
Vision Transformers (ViT) have recently demonstrated success across a myriad of computer vision tasks. However, their elevated computational demands pose significant challenges for real-world deployment. While low-rank approximation stands out as a renowned method to reduce computational loads, efficiently automating the target rank selection in ViT remains a challenge. Drawing from the notable similarity and alignment between the processes of rank selection and One-Shot NAS, we introduce FLORA, an end-to-end automatic framework based on NAS. To overcome the design challenge of supernet posed by vast search space, FLORA employs a low-rank aware candidate filtering strategy. This method adeptly identifies and eliminates underperforming candidates, effectively alleviating potential undertraining and interference among subnetworks. To further enhance the quality of low-rank supernets, we design a low-rank specific training paradigm. First, we propose weight inheritance to construct supernet and enable gradient sharing among low-rank modules. Secondly, we adopt low-rank aware sampling to strategically allocate training resources, taking into account inherited information from pre-trained models. Empirical results underscore FLORA's efficacy. With our method, a more fine-grained rank configuration can be generated automatically and yield up to 33% extra FLOPs reduction compared to a simple uniform configuration. More specific, FLORA-DeiT-B/FLORA-Swin-B can save up to 55%/42% FLOPs almost without performance degradtion. Importantly, FLORA boasts both versatility and orthogonality, offering an extra 21%-26% FLOPs reduction when integrated with leading compression techniques or compact hybrid structures. Our code is publicly available at https://github.com/shadowpa0327/FLORA.
Code (1)
Methods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
One Step Beyond: Feedthrough & Placement-Aware Rectilinear Floorplanner
Floorplanning determines the shapes and locations of modules on a chip canvas and plays a critical role in optimizing the chip's Power, Performance, and Area (PPA) metrics. However, existing floorplanning approaches ofte…
RefLoRA: Refactored Low-Rank Adaptation for Efficient Fine-Tuning of Large Models
Low-Rank Adaptation (LoRA) lowers the computational and memory overhead of fine-tuning large models by updating a low-dimensional subspace of the pre-trained weight matrix. Albeit efficient, LoRA exhibits suboptimal conv…
Natural Language UnderstandingDiffLoRA: Differential Low-Rank Adapters for Large Language Models
Differential Transformer has recently been proposed to improve performance in Transformer models by canceling out noise through a denoiser attention mechanism. In this work, we introduce DiffLoRA, a parameter-efficient a…
parameter-efficient fine-tuningzFLoRA: Zero-Latency Fused Low-Rank Adapters
Large language models (LLMs) are increasingly deployed with task-specific adapters catering to multiple downstream applications. In such a scenario, the additional compute associated with these apparently insignificant n…
AFLoRA: Adaptive Federated Fine-Tuning of Large Language Models with Resource-Aware Low-Rank Adaption
Federated fine-tuning has emerged as a promising approach to adapt foundation models to downstream tasks using decentralized data. However, real-world deployment remains challenging due to the high computational and comm…