Dialogue Response Generation on SaferDialogues
12.4PPLBlenderBot 1
Evaluation results
| Method | Links | ||
|---|---|---|---|
| BlenderBot 1Training=Fine-tuned on curated dialogue datasets2022.05 | 12.4 | 16.1 | |
| R2C2 BlenderBotTraining=Fine-tuned on curated dialogue datasets2022.05 | 13.8 | 16 | |
| OPT-175BParameter count=175B2022.05 | 14.7 | 14.1 | |
| Reddit 2.7BParameter count=2.7B2022.05 | 16.2 | 14 |