Medical Question Answering on MedQuAD (test)
53ROUGE-1DeepSeek R1
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| DeepSeek R1Evaluation Protocol=Fine-tuned2026.02 | 53 | 22 | 21 | 9.8 | |
| LLaMAEvaluation Protocol=Fine-tuned2026.02 | 18 | 12 | 14 | 0.03 | |
| GPT-2Evaluation Protocol=Fine-tuned2026.02 | 16 | 8 | 13 | 0.02 | |
| BioGPTEvaluation Protocol=Zero-Shot2026.02 | 8.4 | 2.8 | 7 | 0.03 |