Multi-modal Understanding on MME v1.0 (test)
1,531.3MME^P ScoreLLaVA-1.5-13B
Evaluation results
| Method | Links | ||
|---|---|---|---|
| LLaVA-1.5-13BModel Size=13B2024.05 | 1,531.3 | 295.4 | |
| CSRBackbone=LLaVA-1.5-13B2024.05 | 1,530.6 | 303.9 | |
| Self-rewardingBackbone=LLaVA-1.5-13B2024.05 | 1,529 | 300.1 | |
| CSRBackbone=LLaVA-1.5-7B2024.05 | 1,524.2 | 367.9 | |
| LLaVA-1.5-7BModel Size=7B2024.05 | 1,510.7 | 348.2 | |
| Self-rewardingBackbone=LLaVA-1.5-7B2024.05 | 1,505.6 | 362.5 | |
| Human-PreferBackbone=LLaVA-1.5-7B2024.05 | 1,490.6 | 335 | |
| RLHF-VBackbone=LLaVA-1.5-7B2024.05 | 1,489.2 | 349.4 | |
| POVIDBackbone=LLaVA-1.5-7B2024.05 | 1,452.8 | 325.3 | |
| VlfeedbackBackbone=LLaVA-1.5-7B2024.05 | 1,432.7 | 321.8 |