Out-of-domain Image Captioning on NoCaps
1.055CIDErVanilla (LLaVA-1.5-7B)
Evaluation Results
| Method | Links | |
|---|---|---|
| Vanilla (LLaVA-1.5-7B)Retained Tokens=5762026.01 | 1.055 | |
| FastVRetained Tokens=1922026.01 | 1.033 | |
| PyramidDropRetained Tokens=1922026.01 | 1.023 | |
| VisionZipRetained Tokens=1922026.01 | 1.023 | |
| ViTCoPRetained Tokens=1922026.01 | 1.019 | |
| SparseVLMRetained Tokens=1922026.01 | 1.01 | |
| ViTCoPRetained Tokens=1282026.01 | 1.008 | |
| FastVRetained Tokens=1282026.01 | 0.995 | |
| VisionZipRetained Tokens=1282026.01 | 0.989 | |
| ViTCoPRetained Tokens=642026.01 | 0.973 | |
| PyramidDropRetained Tokens=1282026.01 | 0.969 | |
| VisionZipRetained Tokens=642026.01 | 0.9 | |
| SparseVLMRetained Tokens=1282026.01 | 0.823 | |
| FastVRetained Tokens=642026.01 | 0.768 | |
| PyramidDropRetained Tokens=642026.01 | 0.627 | |
| SparseVLMRetained Tokens=642026.01 | 0.584 |