ethanrao1Abstract
In the present work we present Training Noise Token (TNT) Pruning for vision transformers. Our method relaxes the discrete token dropping condition to continuous additive noise, providing smooth optimization in training, while retaining discrete dropping computational gains in deployment settings. We provide theoretical connections to Rate-Distortion literature, and empirical evaluations on the ImageNet dataset using ViT and DeiT architectures demonstrating TNT's advantages over previous pruning methods.
Community
We are pleased to share our latest research paper 'Training Noise Token (TNT) Pruning" with the community. In this work, we introduce a novel approach to token pruning for Vision Transformer by relaxing the discrete token dropping condition to continuous additive noise, providing smooth optimization in training, while retaining discrete dropping computational gains in deployment settings. We provide theoretical connections to Rate-Distortion literature, and empirical evaluations on the ImageNet dataset using ViT and DeiT architectures demonstrating TNT's advantages over previous pruning methods. We look forward to hearing your thoughts and feedback!
This is an automated message from the Librarian Bot. I found the following papers similar to this paper.
The following papers were recommended by the Semantic Scholar API
- Token Pruning using a Lightweight Background Aware Vision Transformer (2024)
- Efficient Online Inference of Vision Transformers by Training-Free Tokenization (2024)
- Large Language Models Are Overparameterized Text Encoders (2024)
- LLM Vocabulary Compression for Low-Compute Environments (2024)
- ED-ViT: Splitting Vision Transformer for Distributed Inference on Edge Devices (2024)
- SparseVLM: Visual Token Sparsification for Efficient Vision-Language Model Inference (2024)
- Treat Visual Tokens as Text? But Your MLLM Only Needs Fewer Efforts to See (2024)
Please give a thumbs up to this comment if you found it helpful!
If you want recommendations for any Paper on Hugging Face checkout this Space
You can directly ask Librarian Bot for paper recommendations by tagging it in a comment:
@librarian-bot
recommend
Models citing this paper 0
No model linking this paper
Datasets citing this paper 0
No dataset linking this paper
Spaces citing this paper 0
No Space linking this paper