paper-with-me

홈 › Papers

An Investigation on Hardware-Aware Vision Transformer Scaling

2021-09-29 · Chaojian Li, KyungMin Kim, Bichen Wu, Peizhao Zhang, Hang Zhang, Xiaoliang Dai, Peter Vajda, Yingyan Lin

Vision Transformer (ViT) has demonstrated promising performance in various computer vision tasks, and recently attracted a lot of research attention. Many recent works have focused on proposing new architectures to improve ViT and deploying it into real-world applications. However, little effort has been made to analyze and understand ViT’s architecture design space and its implication of hardware-cost on different devices. In this work, by simply scaling ViT's depth, width, input size, and other basic configurations, we show that a scaled vanilla ViT model without bells and whistles can achieve comparable or superior accuracy-efficiency trade-off than most of the latest ViT variants. Specifically, compared to DeiT-Tiny, our scaled model achieves a $\uparrow1.9\%$ higher ImageNet top-1 accuracy under the same FLOPs and a $\uparrow3.7\%$ better ImageNet top-1 accuracy under the same latency on an NVIDIA Edge GPU TX2. Motivated by this, we further investigate the extracted scaling strategies from the following two aspects: (1) "can these scaling strategies be transferred across different real hardware devices?''; and (2) "can these scaling strategies be transferred to different ViT variants and tasks?''. For (1), our exploration, based on various devices with different resource budgets, indicates that the transferability effectiveness depends on the underlying device together with its corresponding deployment tool; for (2), we validate the effective transferability of the aforementioned scaling strategies obtained from a vanilla ViT model on top of an image classification task to the PiT model, a strong ViT variant targeting efficiency, as well as object detection and video classification tasks. In particular, when transferred to PiT, our scaling strategies lead to a boosted ImageNet top-1 accuracy of from $74.6\%$ to $76.7\%$ ($\uparrow2.1\%$) under the same 0.7G FLOPs; and when transferred to the COCO object detection task, the average precision is boosted by $\uparrow0.7\%$ under a similar throughput on a V100 GPU.

📄 PDF Abstract BibTeX

Code (0)

등록된 구현이 없습니다.

Tasks

GPUimage-classificationImage Classificationobject-detectionObject DetectionVideo Classification

Methods 이 논문이 사용한 방법론

Multi-Head Attention 설명 없음
Attention 설명 없음
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Residual Connection 설명 없음
Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Dropout Dropout is a regularization technique for neural networks that drops a unit (along with connections) at training time with a specified probability $p$ (a common value is…
Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…
Absolute Position Encodings Absolute Position Encodings are a type of position embeddings for [Transformer-based models] where positional encodings are…

Similar Papers 제목 키워드 기반

Quasar-ViT: Hardware-Oriented Quantization-Aware Architecture Search for Vision Transformers

2024-07-25 · Zhengang Li, Alec Lu, Yanyue Xie, Zhenglun Kong 외

Vision transformers (ViTs) have demonstrated their superior accuracy for computer vision tasks compared to convolutional neural networks (CNNs). However, ViT models are often computation-intensive for efficient deploymen…

Quantization

The Energy Consumption of Transformer Fine-Tuning: A Roofline-Inspired Scaling Model

2026-06-22 · Mansour Zoubeirou a Mayaki arxiv

Transformer-based models underpin modern natural language processing but incur rapidly growing computational and energy costs. As training scales in both model size and parallelism, accurately predicting energy consumpti…

Towards Robust Vision Transformer

2021-05-17 · CVPR 2022 1 · Xiaofeng Mao, Gege Qi, Yuefeng Chen, Xiaodan Li 외

Recent advances on Vision Transformer (ViT) and its improved variants have shown that self-attention-based networks surpass traditional Convolutional Neural Networks (CNNs) in most vision tasks. However, existing ViTs fo…

Domain GeneralizationImage ClassificationRobust Design

P$^2$-ViT: Power-of-Two Post-Training Quantization and Acceleration for Fully Quantized Vision Transformer

2024-05-30 · Huihong Shi, Xin Cheng, Wendong Mao, Zhongfeng Wang

Vision Transformers (ViTs) have excelled in computer vision tasks but are memory-consuming and computation-intensive, challenging their deployment on resource-constrained devices. To tackle this limitation, prior works h…

Quantization

Beyond Uniform Scaling: Exploring Depth Heterogeneity in Neural Architectures

2024-02-19 · Akash Guna R. T, Arnav Chavan, Deepak Gupta

Conventional scaling of neural networks typically involves designing a base network and growing different dimensions like width, depth, etc. of the same by some predefined scaling factors. We introduce an automated scali…