Inbox Sorting on PMR-Real (test)
77T-NDCG@10Qwen8B_SFT
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Qwen8B_SFTEvaluation Protocol=UrgentSFT Pairs2026.01 | 77 | 39 | |
| MedGem27B_SFTEvaluation Protocol=UrgentSFT Pairs2026.01 | 75 | 39 | |
| Reward-8B_UrgentEvaluation Protocol=UrgentReward Pairs2026.01 | 71 | 38 | |
| MedGem27BEvaluation Protocol=0-Shot Pairs2026.01 | 70 | 31 | |
| MedGem27B_SFTEvaluation Protocol=Multi-Class2026.01 | 66 | 35 | |
| Reward-4B_UrgentEvaluation Protocol=UrgentReward Pairs2026.01 | 65 | 37 | |
| MedGem27BEvaluation Protocol=Multi-Class2026.01 | 64 | 32 | |
| Qwen32BEvaluation Protocol=Multi-Class2026.01 | 62 | 34 | |
| GPT-OSS*Evaluation Protocol=0-Shot Pairs, Parameters=20B2026.01 | 59 | 32 | |
| Reward-8B_BaseEvaluation Protocol=UrgentReward Pairs2026.01 | 54 | 20 | |
| Qwen8BEvaluation Protocol=0-Shot Pairs2026.01 | 52 | 18 | |
| GPT-OSSEvaluation Protocol=Multi-Class2026.01 | 48 | 23 | |
| Reward-4B_BaseEvaluation Protocol=UrgentReward Pairs2026.01 | 48 | 18 | |
| Qwen32BEvaluation Protocol=0-Shot Pairs2026.01 | 42 | 17 |