Medical Text Generation on MAQA (test)
90.3BERTScoreseverity-aware multi-model framework
Evaluation Results
| Method | Links | |
|---|---|---|
| severity-aware multi-model frameworkTraining Mode=Fine-tuned2026.06 | 90.3 | |
| Dialect-ar-gpt-2021Training Mode=Fine-tuned2026.06 | 86.74 | |
| severity-aware multi-model frameworkTraining Mode=Baseline2026.06 | 86.71 | |
| Aragpt2-BaseTraining Mode=Fine-tuned2026.06 | 85.51 | |
| Aragpt2-MediumTraining Mode=Fine-tuned2026.06 | 85.45 | |
| Qwen3-0.6BTraining Mode=Fine-tuned2026.06 | 84.22 | |
| Bloomz-560MTraining Mode=Fine-tuned2026.06 | 82.81 | |
| Qwen3-0.6BTraining Mode=Baseline2026.06 | 82 | |
| Dialect-ar-gpt-2021Training Mode=Baseline2026.06 | 81.61 | |
| Aragpt2-MediumTraining Mode=Baseline2026.06 | 79.69 | |
| Bloomz-560MTraining Mode=Baseline2026.06 | 78.72 | |
| Aragpt2-BaseTraining Mode=Baseline2026.06 | 72.51 |