Patient Simulation on Ch-PatientSim (test)
0.2054BLEU-1Qwen2.5-72B+MSPRP
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| Qwen2.5-72B+MSPRPbackbone=Qwen2.5-72B, framework=MSPRP2026.01 | 0.2054 | 0.1057 | 0.0659 | 0.045 | 0.2291 | 0.2313 | 0.6575 | |
| Qwen2.5-7B+MSPRPbackbone=Qwen2.5-7B, framework=MSPRP2026.01 | 0.2031 | 0.0956 | 0.0569 | 0.0375 | 0.2196 | 0.2041 | 0.6491 | |
| Qwen2.5-72Bmodel_size=72B2026.01 | 0.2006 | 0.1019 | 0.0634 | 0.0431 | 0.2257 | 0.2256 | 0.6551 | |
| Qwen2.5-7Bmodel_size=7B2026.01 | 0.1828 | 0.0856 | 0.0504 | 0.0329 | 0.2051 | 0.2004 | 0.6437 | |
| GLM-4-9Bmodel_size=9B2026.01 | 0.1824 | 0.0874 | 0.0521 | 0.0346 | 0.2099 | 0.1949 | 0.6478 | |
| Llama-3.1-8Bmodel_size=8B2026.01 | 0.1656 | 0.0724 | 0.0422 | 0.0278 | 0.1839 | 0.1663 | 0.6293 | |
| GPT-4ominivariant=mini2026.01 | 0.1606 | 0.0738 | 0.0434 | 0.0285 | 0.1873 | 0.194 | 0.6383 | |
| Internlm3-8Bmodel_size=8B2026.01 | 0.1552 | 0.0749 | 0.0454 | 0.0308 | 0.1844 | 0.1812 | 0.6299 | |
| DeepSeek-V32026.01 | 0.1348 | 0.0596 | 0.0335 | 0.0214 | 0.1606 | 0.1948 | 0.6266 |