Medical Visual Question Answering on SLAKE-VQA (Accuracy)
81.7AccuracyRform + Rexact + RDTW
Evaluation Results
| Method | Links | |
|---|---|---|
| Rform + Rexact + RDTWReward functions=Rform + Rexact + RDTW2026.04 | 81.7 | |
| Rform + RexactReward functions=Rform + Rexact2026.04 | 77.1 | |
| SFTTraining configuration=SFT2026.04 | 76.1 | |
| RformReward functions=Rform2026.04 | 71.6 |