Multitask Learning on 14 vision tasks
79.5AccuracyTIES
Evaluation results
| Method | Links | |
|---|---|---|
| TIESBackbone=ViT-L/14, Basis size (M)=100%2025.02 | 79.5 | |
| AE (Ours)Backbone=ViT-L/14, Basis size (M)=50%, Merging Method=TIES2025.02 | 75.4 | |
| TABackbone=ViT-B/16, Basis size (M)=100%2025.02 | 70.5 | |
| AE (Ours)Backbone=ViT-B/16, Basis size (M)=50%, Merging Method=TA2025.02 | 67.3 | |
| PretrainedBackbone=ViT-B/162025.02 | 62 |