Image Classification on ImageNet-1K 100% labels (train)
84.2Top-1 AccuracyData2Vec
Evaluation results
| Method | Links | |
|---|---|---|
| Data2VecArchitecture=ViT-B/16, Pretrain Epochs=800, Evaluation protocol=End-to-end fine-tuning2022.04 | 84.2 | |
| MaskFeatArchitecture=ViT-B/16, Evaluation protocol=End-to-end fine-tuning2022.04 | 84 | |
| iBOTArchitecture=ViT-B/16, Pretrain Epochs=800, Evaluation protocol=End-to-end fine-tuning2022.04 | 83.8 | |
| SimMIMArchitecture=ViT-B/16, Evaluation protocol=End-to-end fine-tuning2022.04 | 83.8 | |
| DINOArchitecture=ViT-B/16, Pretrain Epochs=800, Evaluation protocol=End-to-end fine-tuning2022.04 | 83.6 | |
| MAEArchitecture=ViT-B/16, Pretrain Epochs=1600, Evaluation protocol=End-to-end fine-tuning2022.04 | 83.6 | |
| MSNArchitecture=ViT-B/16, Pretrain Epochs=600, Evaluation protocol=End-to-end fine-tuning2022.04 | 83.4 | |
| BEiTArchitecture=ViT-B/16, Pretrain Epochs=800, Evaluation protocol=End-to-end fine-tuning2022.04 | 83.2 |