Dialogue Generation on Norm-grounded Dialogue (Chinese) (test)
79Win Rate vs NormDialQwen-2.5-32B
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Qwen-2.5-32BModel=Qwen-2.5-32B2025.09 | 79 | — | |
| GPT-4o-miniModel=GPT-4o-mini2025.09 | 75 | — | |
| Qwen-2.5-14BModel=Qwen-2.5-14B2025.09 | 71 | — |