Medical Calculation on MedCalc-Bench Original (train)
49.19AccuracyMistral-7B
Evaluation results
| Method | Links | |
|---|---|---|
| Mistral-7BApproach=Supervised fine-tuning, Dataset=Orig (train split), Setting=SFT2026.02 | 49.19 | |
| Llama-2-7BApproach=Supervised fine-tuning, Dataset=Orig (train split), Setting=SFT2026.02 | 45.75 |