Text Generation on DomainBench Geography
0.7082BERTScoreTF-TTCL
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| TF-TTCL2026.04 | 0.7082 | -0.5937 | 8.28 | 31.92 | 12.58 | 24.19 | |
| EATA2026.04 | 0.7033 | -0.6088 | 8.7 | 30.39 | 12.43 | 23.32 | |
| COME2026.04 | 0.6985 | -0.6298 | 7.9 | 29 | 11.58 | 21.61 | |
| TLM2026.04 | 0.698 | -0.6772 | 7.85 | 29.03 | 11.67 | 21.47 | |
| Tent2026.04 | 0.6966 | -0.7273 | 8.57 | 29.59 | 12.69 | 23.68 | |
| Llama-3.1-8B-Instructstatus=Base Model2026.04 | 0.6909 | -0.68 | 6.85 | 27.01 | 10.25 | 19.55 | |
| TF-GRPO2026.04 | 0.6717 | -0.733 | 5.44 | 24.87 | 9.48 | 17.94 |