Papers › DeiT III: Revenge of the ViT

DeiT III: Revenge of the ViT

14 Apr 2022arXiv:2204.07118archive 2025-07-28

Hugo Touvron, Matthieu Cord, Hervé Jégou

A Vision Transformer (ViT) is a simple neural architecture amenable to serve several computer vision tasks. It has limited built-in architectural priors, in contrast to more recent architectures that incorporate priors either about the input data or of specific tasks. Recent works show that ViTs benefit from self-supervised pre-training, in particular BerT-like pre-training like BeiT. In this paper, we revisit the supervised training of ViTs. Our procedure builds upon and simplifies a recipe introduced for training ResNet-50. It includes a new simple data-augmentation procedure with only 3 augmentations, closer to the practice in self-supervised learning. Our evaluations on Image classification (ImageNet-1k with and without pre-training on ImageNet-21k), transfer learning and semantic segmentation show that our procedure outperforms by a large margin previous fully supervised training recipes for ViT. It also reveals that the performance of our ViT trained with supervision is comparable to that of more recent architectures. Our results could serve as better baselines for recent self-supervised approaches demonstrated on ViT.

PaperPDFCode

In Syntology View this paper on Syntology, its page in Syntology's graph. That page lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

facebookresearch/deit officialpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Augmenting Convolutional networks with attention-based aggregation (arXiv:2112.13692; that paper's own run record: 2 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified) and Towards Understanding Why Label Smoothing Degrades Selective Classification and How to Fix It (arXiv:2403.14715; that paper's own run record: 9 ran (of which 2 constructed an object rather than computing a result; 6 with no instrument failure: 1 honoured, 0 violated, 5 with no contract checked; 3 where Syntology's instrument failed) · 1 unverified); nothing checks that they implement this paper's method. report
affjljoo3581/deit3-jax mentioned on GitHubjaxSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale (arXiv:2010.11929; that paper's own run record: 307 ran (of which 165 constructed an object rather than computing a result; 286 with no instrument failure: 8 honoured, 2 violated, 276 with no contract checked; 21 where Syntology's instrument failed) · 112 unverified); nothing checks that they implement this paper's method. report
rwightman/pytorch-image-models mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to AdamP: Slowing Down the Slowdown for Momentum Optimizers on Scale-invariant Weights (arXiv:2006.08217; that paper's own run record: 4 ran (of which 1 constructed an object rather than computing a result; 2 with no instrument failure: 1 honoured, 0 violated, 1 with no contract checked; 2 where Syntology's instrument failed) · 0 unverified); nothing checks that they implement this paper's method. report
2023-MindSpore-1/ms-code-7 mindsporeSyntology: not harvested report
2024-MindSpore-1/Code2 mindsporeSyntology: not harvested report
Mind23-2/MindCode-121 mindsporeSyntology: not harvested report
Mind23-2/MindCode-2 mindsporeSyntology: not harvested report
MindSpore-paper-code-2/code3 mindsporeSyntology: not harvested report
alibaba/EasyCV pytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to End-to-End Object Detection with Transformers (arXiv:2005.12872; that paper's own run record: 70 ran (of which 45 constructed an object rather than computing a result; 62 with no instrument failure: 2 honoured, 1 violated, 59 with no contract checked; 8 where Syntology's instrument failed) · 22 unverified), Training data-efficient image transformers & distillation through attention (arXiv:2012.12877; that paper's own run record: 12 ran (of which 1 constructed an object rather than computing a result; 9 with no instrument failure: 1 honoured, 0 violated, 8 with no contract checked; 3 where Syntology's instrument failed) · 7 unverified), Masked Autoencoders Are Scalable Vision Learners (arXiv:2111.06377; that paper's own run record: 86 ran (of which 40 constructed an object rather than computing a result; 68 with no instrument failure: 8 honoured, 0 violated, 60 with no contract checked; 18 where Syntology's instrument failed) · 51 unverified) and 1 more (every one in this page's JSON twin); nothing checks that they implement this paper's method. report
open-mmlab/mmclassification pytorchApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Reversible Vision Transformers (arXiv:2302.04869; that paper's own run record: 24 ran (of which 5 constructed an object rather than computing a result; 17 with no instrument failure: 0 honoured, 0 violated, 17 with no contract checked; 7 where Syntology's instrument failed) · 5 unverified); nothing checks that they implement this paper's method. report
yangyucheng000/MSpaper mindsporeSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Self-Supervision Can Be a Good Few-Shot Learner (arXiv:2207.09176; that paper's own run record: 5 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 3 where Syntology's instrument failed) · 5 unverified); nothing checks that they implement this paper's method. report
gitlab.com/birder/birder pytorchSyntology: not harvested report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Data AugmentationImage ClassificationSelf-Supervised LearningSemantic SegmentationTransfer Learningimage-classification

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Classification ImageNet ViT-B @384 (DeiT III, 21k) Top 1 Accuracy 86.7% #126 of 1060 Archive leaderboard report
Image Classification ImageNet ViT-L GFLOPs 191.2 #198 of 1060 Archive leaderboard report
Image Classification ImageNet ViT-L Number of params 304.8M #198 of 1060 Archive leaderboard report
Image Classification ImageNet ViT-L Top 1 Accuracy 85.8% #198 of 1060 Archive leaderboard report
Image Classification ImageNet ViT-B @224 (DeiT III, 21k) Top 1 Accuracy 85.7% #207 of 1060 Archive leaderboard report
Image Classification ImageNet ViT-H @224 (DeiT III) Top 1 Accuracy 85.2% #245 of 1060 Archive leaderboard report
Image Classification ImageNet ViT-B @384 (DeiT III) Number of params 87M #270 of 1060 Archive leaderboard report
Image Classification ImageNet ViT-B @384 (DeiT III) Top 1 Accuracy 85.0% #270 of 1060 Archive leaderboard report
Image Classification ImageNet ViT-L @224 (DeiT III) Top 1 Accuracy 84.9% #274 of 1060 Archive leaderboard report
Image Classification ImageNet ViT-B @224 (DeiT III) Top 1 Accuracy 83.8% #386 of 1060 Archive leaderboard report
Image Classification ImageNet ViT-S @384 (DeiT III) GFLOPs 15.5 #428 of 1060 Archive leaderboard report
Image Classification ImageNet ViT-S @384 (DeiT III) Number of params 22M #428 of 1060 Archive leaderboard report
Image Classification ImageNet ViT-S @384 (DeiT III) Top 1 Accuracy 83.4% #428 of 1060 Archive leaderboard report
Image Classification ImageNet ViT-S @224 (DeiT III, 21k) Top 1 Accuracy 83.1% #464 of 1060 Archive leaderboard report
Image Classification ImageNet ViT-S @224 (DeiT III) Top 1 Accuracy 81.4% #643 of 1060 Archive leaderboard report
Image Classification ImageNet ReaL ViT-L @384 (DeiT III, 21k) Number of params 304M #54 of 57 Archive leaderboard report
Image Classification ImageNet ReaL ViT-L @384 (DeiT III, 21k) Top 1 Accuracy 87.7% #54 of 57 Archive leaderboard report
Image Classification ImageNet ReaL ViT-H @224 (DeiT III, 21k) Number of params 632M #55 of 57 Archive leaderboard report
Image Classification ImageNet ReaL ViT-H @224 (DeiT III, 21k) Top 1 Accuracy 87.2% #55 of 57 Archive leaderboard report
Image Classification ImageNet ReaL ViT-L @224 (DeiT III, 21k) Top 1 Accuracy 87.0% #56 of 57 Archive leaderboard report
Semantic Segmentation ADE20K val DeiT-L mIoU 55.6 #29 of 95 Archive leaderboard report
Semantic Segmentation ADE20K val DeiT-B mIoU 54.1 #36 of 95 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Introduced by this paper: 3-Augment

3-AugmentAbsolute Position EncodingsAdamAttentionBPEDense ConnectionsDropoutFixResLabel SmoothingLayer NormalizationLayerScaleLinear LayerMulti-Head AttentionPosition-Wise Feed-Forward LayerResidual ConnectionSoftmaxTransformerVision Transformer

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections