Visual Question Answering on VQA v2 (train)
78.9AccuracyFREE-V-F
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| FREE-V-F#Train Params=1.5B, Backbone=ViT-g, LLM=FlanT5-XL2025.06 | 78.9 | 1.71 | |
| BLIP-2 V-F#Train Params=1.2B, Backbone=ViT-g, LLM=FlanT5-XL2025.06 | 78.8 | 1 | |
| FREE-V-O#Train Params=1.6B, Backbone=ViT-g, LLM=OPT-2.7B2025.06 | 78.7 | 1.77 | |
| BLIP-2 V-O#Train Params=1.2B, Backbone=ViT-g, LLM=OPT-2.7B2025.06 | 78.3 | 1.07 | |
| LeeBLIP#Train Params=1.9B2025.06 | 78.1 | 1.65 | |
| Flamingo80B#Train Params=10.6B2025.06 | 77.9 | — | |
| PABEE-BLIP#Train Params=1.9B2025.06 | 77.4 | 1.39 | |
| OFA#Train Params=930M2025.06 | 75.7 | — | |
| DeeBLIP#Train Params=1.9B2025.06 | 75.4 | 1.52 | |
| BLIP#Train Params=385M2025.06 | 73.9 | — | |
| ALBEF#Train Params=314M2025.06 | 72.3 | — | |
| BLIP-2 ViT-g FlanT5xL#Total params=4.1B, Speed=1.00x, Unsupervised=true2025.06 | 64.9 | — | |
| FREE ViT-g FlanT5xL#Total params=4.5B, Speed=1.45x, Unsupervised=true2025.06 | 64.5 | — | |
| Flamingo80B#Total params=80B, Speed=0.05x, Unsupervised=true2025.06 | 59.1 | — | |
| LeeBLIP#Total params=4.7B, Speed=1.29x, Unsupervised=true2025.06 | 57.7 | — | |
| BLIP-2 ViT-g OPT6.7B#Total params=7.8B, Speed=0.52x, Unsupervised=true2025.06 | 55.9 | — | |
| Flamingo9B#Total params=9.3B, Speed=0.44x, Unsupervised=true2025.06 | 55.7 | — | |
| FREE ViT-g OPT2.7B#Total params=4.3B, Speed=1.51x, Unsupervised=true2025.06 | 55.5 | — | |
| MuE#Total params=4.7B, Speed=1.36x, Unsupervised=true2025.06 | 55.4 | — | |
| BLIP-2 ViT-g OPT2.7B#Total params=3.8B, Speed=1.07x, Unsupervised=true2025.06 | 54.6 | — | |
| Flamingo3B#Total params=3.2B, Speed=1.28x, Unsupervised=true2025.06 | 53.2 | — | |
| PABEE-BLIP#Total params=4.7B, Speed=1.31x, Unsupervised=true2025.06 | 49.6 | — | |
| DeeBLIP#Total params=4.7B, Speed=1.39x, Unsupervised=true2025.06 | 41.3 | — |