Visual Question Answering on MMBench Chinese V1.1
83AccuracyIVC-Prune
Evaluation results
| Method | Links | ||
|---|---|---|---|
| IVC-PruneModels=InternVL 2.5 8B, Average Tokens=50%2026.02 | 83 | 100 | |
| VanillaModels=InternVL 2.5 8B, Average Tokens=100%2026.02 | 82.9 | 100 | |
| IVC-PruneModels=Qwen2.5-VL 7B, Average Tokens=50%2026.02 | 82.6 | 100.2 | |
| PDropModels=InternVL 2.5 8B, Average Tokens=56%2026.02 | 81.5 | 99.1 | |
| VanillaModels=Qwen2.5-VL 7B, Average Tokens=100%2026.02 | 81.4 | 100 | |
| FastVModels=InternVL 2.5 8B, Average Tokens=53%2026.02 | 80.8 | 98.6 | |
| FastVModels=Qwen2.5-VL 7B, Average Tokens=54%2026.02 | 80 | 98.2 | |
| PDropModels=Qwen2.5-VL 7B, Average Tokens=61%2026.02 | 79.5 | 98 | |
| IVC-PruneModels=DeepSeek-VL2 Small-16B, Average Tokens=52%2026.02 | 79 | 100.1 | |
| PDropModels=DeepSeek-VL2 Small-16B, Average Tokens=57%2026.02 | 78.9 | 100 | |
| VanillaModels=DeepSeek-VL2 Small-16B, Average Tokens=100%2026.02 | 78.8 | 100 | |
| FastVModels=DeepSeek-VL2 Small-16B, Average Tokens=54%2026.02 | 77.9 | 98.8 | |
| IVC-PruneModels=LLaVA-v1.5 7B, Average Tokens=28%2026.02 | 42.8 | 101.3 | |
| VanillaModels=LLaVA-v1.5 7B, Average Tokens=100%2026.02 | 41.3 | 100 | |
| FastVModels=LLaVA-v1.5 7B, Average Tokens=30%2026.02 | 40.9 | 99.7 | |
| PDropModels=LLaVA-v1.5 7B, Average Tokens=47%2026.02 | 40.9 | 99.8 |