ViT-G/14

Reported on 3 benchmarks across 1 task · 1 paper · 1 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Computer Vision3 results

Image ClassificationonImageNet V2
Top 1 Accuracy· uses extra data· 2021-06-08
83.33
best: 84.63 (Model soups (BASIC-L))
SOTA
Scaling Vision Transformers arXiv:2106.04560
Image ClassificationonObjectNet
Top-1 Accuracy· uses extra data· 2021-06-08
70.53
best: 82.7 (CoCa)
Scaling Vision Transformers arXiv:2106.04560
Image ClassificationonVTAB-1k
Top-1 Accuracy· uses extra data· 2021-06-08
78.29
best: 79.99 (ALIGN (50 hypers/task))
Scaling Vision Transformers arXiv:2106.04560