Dialogue Response Generation on GROWOVER-DIALOGUE (ALL)
4.7BLEU Score (Month 9)RiLM
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| RiLM2024.06 | 4.7 | — | — | — | — | 4.72 | 4.74 | 4.73 | |
| RaLM-D*2024.06 | 4.68 | — | — | — | — | 4.69 | 4.72 | 4.7 | |
| RaLM2024.06 | 4.44 | — | — | — | — | 4.44 | 4.46 | 4.46 | |
| RaLM-CPcontinuously pre-trained=true2024.06 | 4.43 | — | — | — | — | 4.45 | 4.46 | 4.47 | |
| Self-RAG2024.06 | 2.58 | — | — | — | — | 2.55 | 2.32 | 2.48 | |
| Vanilla2024.06 | 1.12 | — | — | — | — | 1.11 | 1.09 | 1.1 | |
| RaLMRetriever=Contriever, Continuously pre-trained=false2024.06 | — | 2.53 | 2.56 | 2.55 | 2.56 | — | — | — | |
| RaLM-CPRetriever=Contriever, Continuously pre-trained=true2024.06 | — | 2.51 | 2.54 | 2.55 | 2.55 | — | — | — | |
| RaLM-D*Retriever=Contriever, Continuously pre-trained=false2024.06 | — | 2.57 | 2.57 | 2.57 | 2.59 | — | — | — | |
| RiLMRetriever=Contriever, Continuously pre-trained=false2024.06 | — | 2.69 | 2.59 | 2.68 | 2.6 | — | — | — | |
| VanillaRetriever=Contriever, Continuously pre-trained=false2024.06 | — | 1.12 | 1.11 | 1.09 | 1.1 | — | — | — |