Image Classification on ImageNet-1K 1.0 (val) (Top-1 Acc and Training Efficiency)
84.1Top-1 AccuracyBaseline
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| BaselineModel=LV-ViT-M2024.03 | 84.1 | — | 368.7 | |
| ToEModel=LV-ViT-M, r1=0.42024.03 | 84.1 | — | 292.5 | |
| NetworkExpansionModel=LV-ViT-M, Expansion=10 to 202024.03 | 84 | — | 292.7 | |
| BaselineModel=LV-ViT-S2024.03 | 83.3 | — | 237.3 | |
| ToEModel=LV-ViT-S, r1=0.42024.03 | 83.3 | — | 195.3 | |
| NetworkExpansionModel=LV-ViT-S, Expansion=8 to 162024.03 | 82.9 | — | 195.5 | |
| ToEModel=LV-ViT-T, r1=0.42024.03 | 79.4 | — | 113.9 | |
| BaselineModel=LV-ViT-T2024.03 | 79.1 | — | 130.5 | |
| NetworkExpansionModel=LV-ViT-T, Expansion=6 to 122024.03 | 78.8 | — | 114.4 |