Visual Reasoning on NLVR2 loc (val)
77.27AccuracyLXMERT+CATT↑
Evaluation Results
| Method | Links | |
|---|---|---|
| LXMERT+CATT↑GPU Hours=1536 (1080Ti), 1056 (V100), Pre-training Data (Image / Text)=0.18M / 9.18M2021.03 | 77.27 | |
| LXMERT+CATTGPU Hours=960 (1080Ti), 624 (V100), Pre-training Data (Image / Text)=0.18M / 9.18M2021.03 | 76.4 | |
| UNITERGPU Hours=882 (V100), Pre-training Data (Image / Text)=4.20M / 9.58M2021.03 | 75.85 | |
| LXMERTGPU Hours=960 (titan xp), Pre-training Data (Image / Text)=0.18M / 9.18M2021.03 | 74.9 | |
| LXMERT+GPU Hours=816 (1080Ti), Pre-training Data (Image / Text)=0.18M / 9.18M2021.03 | 74.8 |