Task-Oriented Dialogue on MultiWOZ 2.4 (test)
43.8JGACoALM 70B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| CoALM 70BZero-shot=true2025.02 | 43.8 | 69.4 | |
| Llama 3.3 70B InstructZero-shot=true2025.02 | 40.8 | 67.6 | |
| NC-Latent-TODZero-shot=true2025.02 | 39.7 | 68.3 | |
| CoALM 405BZero-shot=true, Note=checkpoint from one completed epoch2025.02 | 38.8 | 66.7 | |
| GPT4o-miniZero-shot=true2025.02 | 38.4 | 69.9 | |
| FNCTODZero-shot=true2025.02 | 37.9 | 44.4 | |
| GPT4oZero-shot=true2025.02 | 36.9 | 75.5 | |
| ToolAceZero-shot=true2025.02 | 34.4 | 18 | |
| CoALM 8BZero-shot=true2025.02 | 30.4 | 51.6 | |
| Mistral-7B-Instruct-v0.3Zero-shot=true2025.02 | 27 | 31.2 | |
| Llama 3.1 8B InstructZero-shot=true2025.02 | 26.3 | 19.9 | |
| Granite-20B-CodeZero-shot=true2025.02 | 21.8 | 10.7 | |
| Hammer 2.0 7BZero-shot=true2025.02 | 21.7 | 23.5 | |
| CodeActAgentZero-shot=true2025.02 | 20.2 | 9.5 | |
| GPT 3.5 TurboZero-shot=true2025.02 | 13.5 | — |