Visual Dialogue on VisDial (test)
45.9MRRBLIP-2 ViT-g FlanT5xL
Evaluation Results
| Method | Links | |
|---|---|---|
| BLIP-2 ViT-g FlanT5xL#Total params=4.1B, Speed=1.00x, Unsupervised=true2025.06 | 45.9 | |
| FREE ViT-g FlanT5xL#Total params=4.5B, Speed=1.45x, Unsupervised=true2025.06 | 45.5 | |
| LeeBLIP#Total params=4.7B, Speed=1.29x, Unsupervised=true2025.06 | 39.1 | |
| BLIP-2 ViT-g OPT6.7B#Total params=7.8B, Speed=0.52x, Unsupervised=true2025.06 | 37.5 | |
| MuE#Total params=4.7B, Speed=1.36x, Unsupervised=true2025.06 | 36.6 | |
| PABEE-BLIP#Total params=4.7B, Speed=1.31x, Unsupervised=true2025.06 | 35.7 | |
| BLIP-2 ViT-g OPT2.7B#Total params=3.8B, Speed=1.07x, Unsupervised=true2025.06 | 34.1 | |
| DeeBLIP#Total params=4.7B, Speed=1.39x, Unsupervised=true2025.06 | 33.4 | |
| FREE ViT-g OPT2.7B#Total params=4.3B, Speed=1.51x, Unsupervised=true2025.06 | 32.3 |