PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
March 18, 20244 citationsOpen Access

Dynamic Tuning Towards Parameter and Inference Efficiency for ViT Adaptation

View Full Paper
WZWangbo ZhaoJTJiasheng TangYHYizeng Han

Key Points

Key points are not available for this paper at this time.

Abstract

Existing parameter-efficient fine-tuning (PEFT) methods have achieved significant success on vision transformers (ViTs) adaptation by improving parameter efficiency. However, the exploration of enhancing inference efficiency during adaptation remains underexplored. This limits the broader application of pre-trained ViT models, especially when the model is computationally extensive. In this paper, we propose Dynamic Tuning (DyT), a novel approach to improve both parameter and inference efficiency for ViT adaptation. Specifically, besides using the lightweight adapter modules, we propose a token dispatcher to distinguish informative tokens from less important ones, allowing the latter to dynamically skip the original block, thereby reducing the redundant computation during inference. Additionally, we explore multiple design variants to find the best practice of DyT. Finally, inspired by the mixture-of-experts (MoE) mechanism, we introduce an enhanced adapter to further boost the adaptation performance. We validate DyT across various tasks, including image/video recognition and semantic segmentation. For instance, DyT achieves comparable or even superior performance compared to existing PEFT methods while evoking only 71%-85% of their FLOPs on the VTAB-1K benchmark.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Zhao et al. (2024) studied this question.

synapsesocial.com/papers/68e73771b6db6435876b14d7https://doi.org/10.48550/arxiv.2403.11808
Ask AI
Helpful
Bookmark
Share
View Full Paper

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Sparse-Tuning: Adapting Vision Transformers with Efficient Fine-tuning and Inference2024
  2. 2Rethinking Low-Rank Adaptation in Vision: Exploring Head-Level Responsiveness across Diverse Tasks2024
  3. 3DA-VPT: Semantic-Guided Visual Prompt Tuning for Vision Transformers2025
  4. 4Parameter-Efficient and Memory-Efficient Tuning for Vision Transformer: A Disentangled Approach2024
  5. 5Parameter Efficient Fine-tuning of Self-supervised ViTs without Catastrophic Forgetting2024