Reward Modeling on UltraMedical (test)
95.8Easy ScorePairRM-LLaMA3-8B
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| PairRM-LLaMA3-8BBackbone=LLaMA3-8B, Type=Pairwise2024.06 | 95.8 | 72.7 | 74.85 | 70.56 | 78.48 | |
| UltraMedRM-8BParameters=8B2024.06 | 94.12 | 73.47 | 77.3 | 77.22 | 80.53 | |
| FsfairX-LLaMA3-RM-v0.1Backbone=LLaMA32024.06 | 92.86 | 70.41 | 73.62 | 67.22 | 76.03 | |
| UltraRM-13bParameters=13B2024.06 | 90.34 | 73.98 | 69.33 | 66.67 | 75.08 | |
| Eurus-RM-7BParameters=7B2024.06 | 89.5 | 72.96 | 73.01 | 68.33 | 75.95 |