https://github.com/JieShibo/PETL-ViT

[ICCV 2023] Binary Adapters, [AAAI 2023] FacT, [Tech report] Convpass
https://github.com/JieShibo/PETL-ViT

parameter-efficient-tuning pytorch transfer-learning visual-prompting

Last synced: over 1 year ago
JSON representation

[ICCV 2023] Binary Adapters, [AAAI 2023] FacT, [Tech report] Convpass

• Host: GitHub
• URL: https://github.com/JieShibo/PETL-ViT
• Owner: JieShibo
• License: mit
• Created: 2022-07-08T14:31:22.000Z (about 4 years ago)
• Default Branch: main
• Last Pushed: 2023-08-01T03:12:58.000Z (almost 3 years ago)
• Last Synced: 2024-08-02T15:33:27.599Z (almost 2 years ago)
• Topics: parameter-efficient-tuning, pytorch, transfer-learning, visual-prompting
• Language: Python
• Homepage:
• Size: 18.5 MB
• Stars: 164
• Watchers: 2
• Forks: 7
• Open Issues: 2
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-parameter-efficient-fine-tuning-for-foundation-models - [code

README

          
# Parameter-Efficient Transfer Learning on Vision Transformers

Some exploration of *Parameter-Efficient Transfer Learning* in vision.

+ **Revisiting the Parameter Efficiency of Adapters from the Perspective of Precision Redundancy**

    + \[ICCV 2023\] [![Arxiv link](https://img.shields.io/static/v1?label=arXiv&message=Binary_Adapter&color=red&logo=arxiv)](https://arxiv.org/abs/2307.16867)

[![Github link](https://img.shields.io/static/v1?label=GitHub&message=Binary_Adapter&color=black&logo=github)](https://github.com/JieShibo/PETL-ViT/tree/main/binary_adapter)

+ **FacT: Factor-Tuning for Lightweight Adaptation on Vision Transformer**

    + \[AAAI 2023 Oral\] [![Arxiv link](https://img.shields.io/static/v1?label=arXiv&message=FacT&color=red&logo=arxiv)](https://arxiv.org/abs/2212.03145)

[![Github link](https://img.shields.io/static/v1?label=GitHub&message=FacT&color=black&logo=github)](https://github.com/JieShibo/PETL-ViT/tree/main/FacT)

+ **Convolutional Bypasses Are Better Vision Transformer Adapters**

    + [Tech Report 2022] [![Arxiv link](https://img.shields.io/static/v1?label=arXiv&message=Convpass&color=red&logo=arxiv)](https://arxiv.org/abs/2207.07039)

[![Github link](https://img.shields.io/static/v1?label=GitHub&message=Convpass&color=black&logo=github)](https://github.com/JieShibo/PETL-ViT/tree/main/convpass)