Response Quality Evaluation on Controversial Topics (test)
1.17Win ScoreDEBATUNE-13B
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| DEBATUNE-13BParameters=13B2024.02 | 1.17 | 43 | 101 | 16 | |
| DEBATUNE-7BParameters=7B, Role=baseline2024.02 | 1 | — | — | — | |
| WizardLM 13B V1.2Parameters=13B, Version=V1.22024.02 | 0.79 | 14 | 99 | 47 | |
| Zephyr 7B BetaParameters=7B, Version=Beta2024.02 | 0.67 | 12 | 84 | 64 | |
| Vicuna 13B v1.5Parameters=13B, Version=v1.52024.02 | 0.29 | 5 | 36 | 119 | |
| Zephyr 7B AlphaParameters=7B, Version=Alpha2024.02 | 0.27 | 7 | 29 | 124 | |
| LLaMA2 Chat 13BParameters=13B2024.02 | 0.2 | 3 | 26 | 131 | |
| Vicuna 7B v1.5Parameters=7B, Version=v1.52024.02 | 0.19 | 6 | 19 | 135 | |
| LLaMA2 Chat 7BParameters=7B2024.02 | 0.12 | 1 | 17 | 142 | |
| WizardLM 7BParameters=7B2024.02 | 0.11 | 3 | 12 | 145 | |
| Alpaca 7BParameters=7B2024.02 | 0.03 | 2 | 1 | 157 |