Dialogue Evaluation on ConvAI2 (C2)
10.2PerplexityBlenderBot 1
Evaluation results
| Method | Links | ||
|---|---|---|---|
| BlenderBot 1Evaluation Mode=Supervised2022.05 | 10.2 | 0.183 | |
| R2C2 BlenderBotEvaluation Mode=Supervised2022.05 | 10.5 | 0.205 | |
| OPT-175BEvaluation Mode=Unsupervised2022.05 | 10.8 | 0.185 | |
| Reddit 2.7BEvaluation Mode=Unsupervised2022.05 | 18.9 | 0.126 |