Accuracy on VQA v2
78.53AccuracyLLaVA-1.5-7B
Evaluation results
| Method | Links | |
|---|---|---|
| LLaVA-1.5-7BToken Retention Rate=576, Base Model=LLaVA-1.5-7B2025.06 | 78.53 | |
| VisionDropToken Retention Rate=192, Base Model=LLaVA-1.5-7B2025.06 | 77.28 | |
| SparseVLM-v1.5Token Retention Rate=192, Base Model=LLaVA-1.5-7B2025.06 | 77.01 | |
| VisPrunerToken Retention Rate=192, Base Model=LLaVA-1.5-7B2025.06 | 76.96 | |
| VisionZipToken Retention Rate=192, Base Model=LLaVA-1.5-7B2025.06 | 76.78 | |
| VisionDropToken Retention Rate=128, Base Model=LLaVA-1.5-7B2025.06 | 76.24 | |
| SparseVLM-v1.5Token Retention Rate=128, Base Model=LLaVA-1.5-7B2025.06 | 76.21 | |
| VisPrunerToken Retention Rate=128, Base Model=LLaVA-1.5-7B2025.06 | 75.74 | |
| VisionZipToken Retention Rate=128, Base Model=LLaVA-1.5-7B2025.06 | 75.6 | |
| PDropToken Retention Rate=192, Base Model=LLaVA-1.5-7B2025.06 | 75.57 | |
| PDropToken Retention Rate=128, Base Model=LLaVA-1.5-7B2025.06 | 74.21 | |
| VisionDropToken Retention Rate=64, Base Model=LLaVA-1.5-7B2025.06 | 73.16 | |
| VisPrunerToken Retention Rate=64, Base Model=LLaVA-1.5-7B2025.06 | 72.64 | |
| VisionZipToken Retention Rate=64, Base Model=LLaVA-1.5-7B2025.06 | 72.42 | |
| FastVToken Retention Rate=192, Base Model=LLaVA-1.5-7B2025.06 | 70.51 | |
| SparseVLM-v1.5Token Retention Rate=64, Base Model=LLaVA-1.5-7B2025.06 | 70.24 | |
| VisionDropToken Retention Rate=32, Base Model=LLaVA-1.5-7B2025.06 | 68.55 | |
| VisPrunerToken Retention Rate=32, Base Model=LLaVA-1.5-7B2025.06 | 67.41 | |
| VisionZipToken Retention Rate=32, Base Model=LLaVA-1.5-7B2025.06 | 67.12 | |
| FastVToken Retention Rate=128, Base Model=LLaVA-1.5-7B2025.06 | 66.61 | |
| PDropToken Retention Rate=64, Base Model=LLaVA-1.5-7B2025.06 | 63.04 | |
| FastVToken Retention Rate=64, Base Model=LLaVA-1.5-7B2025.06 | 61.08 |