Human Evaluation of Dialogue on Persona-Chat 1.0 (test)
4.31FluencyHuman
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| HumanProfile=Self2018.01 | 4.31 | 4.25 | 4.36 | 95 | |
| KV Profile MemoryModel Category=Ranking, Profile=Self2018.01 | 3.97 | 3.5 | 3.44 | 81 | |
| KV MemoryModel Category=Ranking, Profile=None2018.01 | 3.81 | 3.88 | 3.36 | 59 | |
| Twitter LMProfile=None2018.01 | 3.21 | 1.75 | 1.95 | 57 | |
| Seq2SeqModel Category=Generative, Profile=None2018.01 | 3.17 | 3.18 | 2.98 | 51 | |
| Profile MemoryModel Category=Generative, Profile=Self2018.01 | 3.08 | 3.13 | 3.14 | 72 | |
| OpenSubtitles 2018 LMProfile=None2018.01 | 2.85 | 2.13 | 2.15 | 35 | |
| OpenSubtitles 2009 LMProfile=None2018.01 | 2.25 | 2.12 | 1.96 | 38 | |
| OpenSubtitles 2009 KV MemoryProfile=None2018.01 | 2.14 | 2.22 | 2.06 | 42 |