Synthetic Dialogue Generation Evaluation on MTS-dialog (test) benchmark leaderboard · SOTA2 Research

Synthetic Dialogue Generation Evaluation on MTS-dialog (test)

52.7ROUGE-1 (Source-Hypothesis)GPT4

Evaluation Results

Method	Links
GPT42023.10		52.7	53.29	20.2	50.81	71.46	45.69	55.17	25.7	49.63	0.019	0.009	0.019
ChatGPT2023.10		43.73	48.56	16.74	46.36	67.54	35.75	46.23	19.72	40.54	0.017	0.006	0.017
NoteChat2023.10		37.24	56.48	19.74	53.41	48.23	51.23	49.68	20.83	36.04	0.014	0.007	0.014
Human2023.10		35.29	—	—	—	—	—	—	14.38	32.89	—	—	—