ConcatPlexer: Additional Dim1 Batching for Faster ViTs
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2023
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910311743750144 |
|---|---|
| author | Han, Donghoon Seo, Seunghyeon Jeon, Donghyeon Jang, Jiho Kong, Chaerin Kwak, Nojun |
| author_facet | Han, Donghoon Seo, Seunghyeon Jeon, Donghyeon Jang, Jiho Kong, Chaerin Kwak, Nojun |
| contents | Transformers have demonstrated tremendous success not only in the natural language processing (NLP) domain but also the field of computer vision, igniting various creative approaches and applications. Yet, the superior performance and modeling flexibility of transformers came with a severe increase in computation costs, and hence several works have proposed methods to reduce this burden. Inspired by a cost-cutting method originally proposed for language models, Data Multiplexing (DataMUX), we propose a novel approach for efficient visual recognition that employs additional dim1 batching (i.e., concatenation) that greatly improves the throughput with little compromise in the accuracy. We first introduce a naive adaptation of DataMux for vision models, Image Multiplexer, and devise novel components to overcome its weaknesses, rendering our final model, ConcatPlexer, at the sweet spot between inference speed and accuracy. The ConcatPlexer was trained on ImageNet1K and CIFAR100 dataset and it achieved 23.5% less GFLOPs than ViT-B/16 with 69.5% and 83.4% validation accuracy, respectively. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2308_11199 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | ConcatPlexer: Additional Dim1 Batching for Faster ViTs Han, Donghoon Seo, Seunghyeon Jeon, Donghyeon Jang, Jiho Kong, Chaerin Kwak, Nojun Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning Transformers have demonstrated tremendous success not only in the natural language processing (NLP) domain but also the field of computer vision, igniting various creative approaches and applications. Yet, the superior performance and modeling flexibility of transformers came with a severe increase in computation costs, and hence several works have proposed methods to reduce this burden. Inspired by a cost-cutting method originally proposed for language models, Data Multiplexing (DataMUX), we propose a novel approach for efficient visual recognition that employs additional dim1 batching (i.e., concatenation) that greatly improves the throughput with little compromise in the accuracy. We first introduce a naive adaptation of DataMux for vision models, Image Multiplexer, and devise novel components to overcome its weaknesses, rendering our final model, ConcatPlexer, at the sweet spot between inference speed and accuracy. The ConcatPlexer was trained on ImageNet1K and CIFAR100 dataset and it achieved 23.5% less GFLOPs than ViT-B/16 with 69.5% and 83.4% validation accuracy, respectively. |
| title | ConcatPlexer: Additional Dim1 Batching for Faster ViTs |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2308.11199 |