Captioning on GovLA-10K
53.32BLEU-1GovLA-Reasoner
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| GovLA-ReasonerLLM Component=Qwen3-4B2026.01 | 53.32 | 37.1 | 26.98 | 20.31 | 26.63 | 37.97 | 20.31 | |
| Qwen3-VL-4BVersion=3, Size=4B2026.01 | 45.77 | 27.72 | 17.41 | 11.36 | 23.25 | 28.92 | 10.22 | |
| Qwen3-VL-8BVersion=3, Size=8B2026.01 | 40.88 | 25.64 | 16.54 | 10.97 | 21.73 | 29.44 | 10.21 | |
| Qwen2.5-VL-3BVersion=2.5, Size=3B2026.01 | 37.17 | 21.25 | 12.71 | 8.04 | 19.21 | 25.01 | 5.26 | |
| LLaVA-OneVision-1.5-4BVersion=1.5, Size=4B2026.01 | 36.27 | 21.24 | 12.56 | 7.61 | 19.1 | 25.36 | 4.84 | |
| Qwen2.5-VL-7BVersion=2.5, Size=7B2026.01 | 36.15 | 21.51 | 13.2 | 8.65 | 19.54 | 25.63 | 5.07 | |
| InternVL3.5-8BVersion=3.5, Size=8B2026.01 | 34.56 | 18.82 | 10.06 | 5.64 | 18.28 | 22.44 | 3.01 | |
| InternVL3-8BVersion=3, Size=8B2026.01 | 31.72 | 17.27 | 9.14 | 5.14 | 17.68 | 22.39 | 2.72 | |
| InternVL3.5-4BVersion=3.5, Size=4B2026.01 | 31.01 | 17.01 | 9.34 | 5.28 | 17.33 | 22.61 | 2.71 | |
| LLaVA-OneVision-1.5-8BVersion=1.5, Size=8B2026.01 | 30.61 | 18.28 | 10.86 | 6.78 | 17.25 | 24.82 | 2.69 |