Medical Report Generation on Merlin (test)
35.64RaTE ScoreAstra
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Astra2026.05 | 35.64 | 37.65 | 41.07 | 37.5 | 46.84 | 40.76 | |
| R2GenGPT2026.05 | 32.29 | 29.65 | 31.4 | 28.82 | 38.51 | 32.1 | |
| LLaVA2026.05 | 32.04 | 26.4 | 30.09 | 27.49 | 39.28 | 30.82 | |
| HuluMed-14BParameters=14B2026.05 | 21.71 | 13.89 | 14.16 | 13.82 | 12.4 | 13.57 | |
| Gemini-32026.05 | 21.54 | 9.75 | 12.99 | 11.7 | 17.43 | 12.96 | |
| HuluMed-32BParameters=32B2026.05 | 21.15 | 12.65 | 19.68 | 14.45 | 12.03 | 14.42 | |
| Qwen3-VL-8BParameters=8B2026.05 | 20.07 | 10.01 | 9.54 | 7.35 | 5.43 | 8.08 | |
| HuluMed-7BParameters=7B2026.05 | 19.63 | 9.97 | 11.53 | 7.17 | 7.44 | 9.03 | |
| Lingshu-32BParameters=32B2026.05 | 19.55 | 4.42 | 8.8 | 6.12 | 4.68 | 6 | |
| M3D2026.05 | 18.91 | 9.12 | 10.64 | 6.35 | 4.76 | 7.72 | |
| MedGemma-4BParameters=4B2026.05 | 17.75 | 3.92 | 6.44 | 3.98 | 2.48 | 4.21 | |
| Lingshu-7BParameters=7B2026.05 | 17.67 | 3.71 | 6.35 | 3.82 | 4.13 | 4.5 | |
| MedGemma-27BParameters=27B2026.05 | 17.19 | 1.95 | 5.7 | 3.22 | 2.18 | 3.26 | |
| RadFM2026.05 | 16.87 | 4.31 | 4.42 | 2.25 | 1.53 | 3.13 |