Task-Focused Dialogue on Multiwoz (G-Eval)
3.453G-Eval ScoreGPT-5.2
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-5.22026.01 | 3.453 | |
| GOPO-Qwen3-14BNumber of parameters=14B2026.01 | 3.447 | |
| Gemini-2.52026.01 | 3.407 | |
| GLM-4.72026.01 | 3.381 | |
| DeepSeek-R12026.01 | 3.368 | |
| Qwen-235BNumber of parameters=235B2026.01 | 3.353 |