Counseling Dialogue Generation on CounselBench Eval (test)
4.29Overall ScoreLlama3-G2C
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Llama3-G2CTraining=fine-tuned, Grounding=Graph2Counsel2026.04 | 4.29 | 4.26 | 4.3 | 4 | 3.98 | 1 | |
| Llama3-MAGTraining=fine-tuned2026.04 | 4.12 | 4.09 | 4.12 | 8 | 3.98 | 1 | |
| Llama3-SQPTraining=fine-tuned2026.04 | 4.07 | 4.06 | 4.08 | 7 | 3.98 | 1 | |
| CAMELTraining=fine-tuned2026.04 | 3.9 | 3.94 | 3.91 | 4 | 3.97 | 1 |