[Preprint] "Chasing Sparsity in Vision Transformers: An End-to-End Exploration" by Tianlong Chen, Yu Cheng, Zhe Gan, Lu Yuan, Lei Zhang, Zhangyang Wang

Overview

Chasing Sparsity in Vision Transformers: An End-to-End Exploration

License: MIT

Codes for [Preprint] Chasing Sparsity in Vision Transformers: An End-to-End Exploration.

Tianlong Chen, Yu Cheng, Zhe Gan, Lu Yuan, Lei Zhang, Zhangyang Wang

Overall Results

Extensive results on ImageNet with diverse ViT backbones validate the effectiveness of our proposals which obtain significantly reduced computational cost and almost unimpaired generalization. Perhaps most surprisingly, we find that the proposed sparse (co-)training can even improve the ViT accuracy rather than compromising it, making sparsity a tantalizing “free lunch”. For example, our sparsified DeiT-Small at (5%, 50%) sparsity for (data, architecture), improves 0.28% top-1 accuracy, and meanwhile enjoys 49.32% FLOPs and 4.40% running time savings.

Proposed Framework of SViTE

Implementations of SViTE

Set Environment

conda create -n vit python=3.6

pip install torch==1.7.1+cu101 torchvision==0.8.2+cu101 torchaudio==0.7.2 -f https://download.pytorch.org/whl/torch_stable.html

pip install tqdm scipy timm

git clone https://github.com/NVIDIA/apex

cd apex

pip install -v --disable-pip-version-check --no-cache-dir --global-option="--cpp_ext" --global-option="--cuda_ext" ./

pip install -v --disable-pip-version-check --no-cache-dir ./

Cmd

Command for unstructured sparsity, i.e., SViTE.

  • SViTE-Small
bash cmd/ vm/0426/vm1.sh 0,1,2,3,4,5,6,7

Details

CUDA_VISIBLE_DEVICES=$1 \
python -m torch.distributed.launch \
    --nproc_per_node=8 \
    --use_env main.py \
    --model deit_small_patch16_224 \
    --epochs 600 \
    --batch-size 64 \
    --data-path ../../imagenet \
    --output_dir ./small_dst_uns_0426_vm1 \
    --dist_url tcp://127.0.0.1:23305 \
    --sparse_init fixed_ERK \
    --density 0.4 \
    --update_frequency 15000 \
    --growth gradient \
    --death magnitude \
    --redistribution none
  • SViTE-Base
bash cmd/ vm/0426/vm3.sh 0,1,2,3,4,5,6,7

Details

CUDA_VISIBLE_DEVICES=$1 \
python -m torch.distributed.launch \
    --nproc_per_node=8 \
    --use_env main.py \
    --model deit_base_patch16_224 \
    --epochs 600 \
    --batch-size 128 \
    --data-path ../../imagenet \
    --output_dir ./base_dst_uns_0426_vm3 \
    --dist_url tcp://127.0.0.1:23305 \
    --sparse_init fixed_ERK \
    --density 0.4 \
    --update_frequency 7000 \
    --growth gradient \
    --death magnitude \
    --redistribution none

Remark. More commands can be found under the "cmd" folder.

Command for structured sparsity is comming soon!

Pre-traiend SViTE Models.

  1. SViTE-Base with 40% structural sparsity ACC=82.22

https://www.dropbox.com/s/ix7mmduvf0wlc4b/deit_base_structure_40_82.22.pth?dl=0

  1. SViTE-Base with 40% unstructured sparsity ACC=81.56

https://www.dropbox.com/s/vltm4piwn9cwsop/deit_base_unstructure_40_81.56.pth?dl=0

  1. SViTE-Small with 50% unstructued sparsity and 5% data sparisity ACC=80.18

https://www.dropbox.com/s/kofps21g857wlbt/deit_small_unstructure_50_sparseinput_0.95_80.18.pth?dl=0

  1. SViTE-Small with 50% unstructured sparsity and 10% data sparsity ACC=79.91

https://www.dropbox.com/s/bdhpc6nfrwahcuc/deit_small_unstructure_50_sparseinput_0.90_79.91.pth?dl=0

Citation

@misc{chen2021chasing,
      title={Chasing Sparsity in Vision Transformers:An End-to-End Exploration}, 
      author={Tianlong Chen and Yu Cheng and Zhe Gan and Lu Yuan and Lei Zhang and Zhangyang Wang},
      year={2021},
      eprint={2106.04533},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}

Acknowledge Related Repos

ViT : https://github.com/jeonsworld/ViT-pytorch

ViT : https://github.com/google-research/vision_transformer

Rig : https://github.com/google-research/rigl

DeiT: https://github.com/facebookresearch/deit

Owner
VITA
Visual Informatics Group @ University of Texas at Austin
VITA
A hyperparameter optimization framework

Optuna: A hyperparameter optimization framework Website | Docs | Install Guide | Tutorial Optuna is an automatic hyperparameter optimization software

7.4k Jan 04, 2023
Code for ICLR 2020 paper "VL-BERT: Pre-training of Generic Visual-Linguistic Representations".

VL-BERT By Weijie Su, Xizhou Zhu, Yue Cao, Bin Li, Lewei Lu, Furu Wei, Jifeng Dai. This repository is an official implementation of the paper VL-BERT:

Weijie Su 698 Dec 18, 2022
Causal Imitative Model for Autonomous Driving

Causal Imitative Model for Autonomous Driving Mohammad Reza Samsami, Mohammadhossein Bahari, Saber Salehkaleybar, Alexandre Alahi. arXiv 2021. [Projec

VITA lab at EPFL 8 Oct 04, 2022
ENet: A Deep Neural Network Architecture for Real-Time Semantic Segmentation

ENet in Caffe Execution times and hardware requirements Network 1024x512 1280x720 Parameters Model size (fp32) ENet 20.4 ms 32.9 ms 0.36 M 1.5 MB SegN

Timo Sämann 561 Jan 04, 2023
For holding anime-related object classification and detection models

Animesion An end-to-end framework for anime-related object classification, detection, segmentation, and other models. Update: 01/22/2020. Due to time-

Edwin Arkel Rios 72 Nov 30, 2022
OpenFace – a state-of-the art tool intended for facial landmark detection, head pose estimation, facial action unit recognition, and eye-gaze estimation.

OpenFace 2.2.0: a facial behavior analysis toolkit Over the past few years, there has been an increased interest in automatic facial behavior analysis

Tadas Baltrusaitis 5.8k Dec 31, 2022
This provides the R code and data to replicate results in "The USS Trustee’s risky strategy"

USSBriefs2021 This provides the R code and data to replicate results in "The USS Trustee’s risky strategy" by Neil M Davies, Jackie Grant and Chin Yan

1 Oct 30, 2021
Fast RFC3339 compliant Python date-time library

udatetime: Fast RFC3339 compliant date-time library Handling date-times is a painful act because of the sheer endless amount of formats used by people

Simon Pirschel 235 Oct 25, 2022
This repository contains code accompanying the paper "An End-to-End Chinese Text Normalization Model based on Rule-Guided Flat-Lattice Transformer"

FlatTN This repository contains code accompanying the paper "An End-to-End Chinese Text Normalization Model based on Rule-Guided Flat-Lattice Transfor

THUHCSI 74 Nov 28, 2022
Code accompanying the NeurIPS 2021 paper "Generating High-Quality Explanations for Navigation in Partially-Revealed Environments"

Generating High-Quality Explanations for Navigation in Partially-Revealed Environments This work presents an approach to explainable navigation under

RAIL Group @ George Mason University 1 Oct 28, 2022
PointCNN: Convolution On X-Transformed Points (NeurIPS 2018)

PointCNN: Convolution On X-Transformed Points Created by Yangyan Li, Rui Bu, Mingchao Sun, Wei Wu, Xinhan Di, and Baoquan Chen. Introduction PointCNN

Yangyan Li 1.3k Dec 21, 2022
Talk covering the features of skorch

Skorch Talk Skorch - A Union of Scikit-learn and PyTorch Presentation The slides can be downloaded at: download link. Google Colab Part One - MNIST Pa

Thomas J. Fan 3 Oct 20, 2020
DeepLM: Large-scale Nonlinear Least Squares on Deep Learning Frameworks using Stochastic Domain Decomposition (CVPR 2021)

DeepLM DeepLM: Large-scale Nonlinear Least Squares on Deep Learning Frameworks using Stochastic Domain Decomposition (CVPR 2021) Run Please install th

Jingwei Huang 130 Dec 02, 2022
Code release for BlockGAN: Learning 3D Object-aware Scene Representations from Unlabelled Images

BlockGAN Code release for BlockGAN: Learning 3D Object-aware Scene Representations from Unlabelled Images BlockGAN: Learning 3D Object-aware Scene Rep

41 May 18, 2022
Official Pytorch Implementation for Splicing ViT Features for Semantic Appearance Transfer presenting Splice

Splicing ViT Features for Semantic Appearance Transfer [Project Page] Splice is a method for semantic appearance transfer, as described in Splicing Vi

Omer Bar Tal 253 Jan 06, 2023
68 keypoint annotations for COFW test data

68 keypoint annotations for COFW test data This repository contains manually annotated 68 keypoints for COFW test data (original annotation of CFOW da

31 Dec 06, 2022
EdMIPS: Rethinking Differentiable Search for Mixed-Precision Neural Networks

EdMIPS is an efficient algorithm to search the optimal mixed-precision neural network directly without proxy task on ImageNet given computation budgets. It can be applied to many popular network arch

Zhaowei Cai 47 Dec 30, 2022
Projecting interval uncertainty through the discrete Fourier transform

Projecting interval uncertainty through the discrete Fourier transform This repo

1 Mar 02, 2022
The materials used in the SaxonJS tutorial presented at Declarative Amsterdam, 2021

SaxonJS-Tutorial-2021, version 1.0.4 Last updated on 4 November, 2021. Table of contents Background Prerequisites Starting a web server Running a Java

Saxonica 11 Oct 23, 2022