Short-form Question Answering on Musique (test)
13.67EMVSPO + PRS_short
Evaluation Results
| Method | Links | |
|---|---|---|
| VSPO + PRS_shortBackbone=Qwen2.5-3B-Instruct2025.12 | 13.67 | |
| GRPO + PRS_shortBackbone=Qwen2.5-3B-Instruct2025.12 | 12.45 | |
| PPO + PRS_shortBackbone=Qwen2.5-3B-Instruct2025.12 | 12 | |
| SFTBackbone=Qwen2.5-3B-Instruct2025.12 | 4.4 | |
| UntrainedBackbone=Qwen2.5-3B-Instruct2025.12 | 0.82 |