Image Editing on IEdit
11.34AccuracyTrimTokenator-LC
Evaluation Results
| Method | Links | |
|---|---|---|
| TrimTokenator-LCBase Model=InternVL-Chat-ViT-6B-Vicuna-13B2025.12 | 11.34 | |
| DivPruneBase Model=InternVL-Chat-ViT-6B-Vicuna-13B2025.12 | 10.77 | |
| Original ModelBase Model=InternVL-Chat-ViT-6B-Vicuna-13B2025.12 | 10.74 | |
| TrimTokenatorBase Model=InternVL-Chat-ViT-6B-Vicuna-13B2025.12 | 10.61 | |
| SparseVLMBase Model=InternVL-Chat-ViT-6B-Vicuna-13B2025.12 | 10.32 | |
| VisionZipBase Model=InternVL-Chat-ViT-6B-Vicuna-13B2025.12 | 10.28 | |
| CATPBase Model=InternVL-Chat-ViT-6B-Vicuna-13B2025.12 | 10.12 | |
| FastVBase Model=InternVL-Chat-ViT-6B-Vicuna-13B2025.12 | 9.82 | |
| VTWBase Model=InternVL-Chat-ViT-6B-Vicuna-13B2025.12 | 9.75 | |
| Original ModelBackbone=LLaVA-1.5-13B, Retention Ratio=1.02025.12 | 9 | |
| TrimTokenator-LCBackbone=LLaVA-1.5-13B, Retention Ratio=0.22025.12 | 8.35 | |
| DivPruneBackbone=LLaVA-1.5-13B, Retention Ratio=0.22025.12 | 8.3 | |
| TrimTokenatorBackbone=LLaVA-1.5-13B, Retention Ratio=0.22025.12 | 8.15 | |
| SparseVLMBackbone=LLaVA-1.5-13B, Retention Ratio=0.22025.12 | 7.96 | |
| Original ModelBackbone=Yi-VL-6B, Retention Ratio=1.02025.12 | 7.9 | |
| VisionZipBackbone=LLaVA-1.5-13B, Retention Ratio=0.22025.12 | 7.82 | |
| VTWBackbone=LLaVA-1.5-13B, Retention Ratio=0.22025.12 | 7.61 | |
| TrimTokenator-LCBackbone=Yi-VL-6B, Retention Ratio=0.22025.12 | 7.59 | |
| CATPBackbone=LLaVA-1.5-13B, Retention Ratio=0.22025.12 | 7.57 | |
| FastVBackbone=LLaVA-1.5-13B, Retention Ratio=0.22025.12 | 7.48 | |
| DivPruneBackbone=Yi-VL-6B, Retention Ratio=0.22025.12 | 7.45 | |
| TrimTokenatorBackbone=Yi-VL-6B, Retention Ratio=0.22025.12 | 7.4 | |
| VisionZipBackbone=Yi-VL-6B, Retention Ratio=0.22025.12 | 7.29 | |
| VTWBackbone=Yi-VL-6B, Retention Ratio=0.22025.12 | 7.25 | |
| SparseVLMBackbone=Yi-VL-6B, Retention Ratio=0.22025.12 | 7.18 | |
| CATPBackbone=Yi-VL-6B, Retention Ratio=0.22025.12 | 7.15 | |
| FastVBackbone=Yi-VL-6B, Retention Ratio=0.22025.12 | 7.12 | |
| Original ModelBackbone=LLaVA-1.5-7B, Retention Ratio=1.02025.12 | 5.36 | |
| TrimTokenator-LCBackbone=LLaVA-1.5-7B, Retention Ratio=0.22025.12 | 5.15 | |
| SparseVLMBackbone=LLaVA-1.5-7B, Retention Ratio=0.22025.12 | 5.03 | |
| DivPruneBackbone=LLaVA-1.5-7B, Retention Ratio=0.22025.12 | 4.61 | |
| TrimTokenatorBackbone=LLaVA-1.5-7B, Retention Ratio=0.22025.12 | 4.52 | |
| VisionZipBackbone=LLaVA-1.5-7B, Retention Ratio=0.22025.12 | 4.43 | |
| VTWBackbone=LLaVA-1.5-7B, Retention Ratio=0.22025.12 | 4.37 | |
| FastVBackbone=LLaVA-1.5-7B, Retention Ratio=0.22025.12 | 4.26 | |
| CATPBackbone=LLaVA-1.5-7B, Retention Ratio=0.22025.12 | 4.21 |