Medical Search Relevance Assessment on LocalQSMed Success-only v1.0
91.9AccuracyQwen3-8B
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen3-8BLoss Rate=0.00%2025.12 | 91.9 | |
| Qwen3-4BLoss Rate=0.00%2025.12 | 91.68 | |
| DeepSeek-R1-Distill-Qwen-14BLoss Rate=0.00%2025.12 | 91.25 | |
| Qwen3-14BLoss Rate=0.02%2025.12 | 90.92 | |
| Llama-3.3-70B-InstructLoss Rate=46.86%2025.12 | 90.37 | |
| DeepSeek-R1-Distill-Llama-70BLoss Rate=43.68%2025.12 | 90.04 | |
| Qwen3-32BLoss Rate=0.00%2025.12 | 89.93 | |
| Llama-3.1-8BLoss Rate=63.55%2025.12 | 88.73 | |
| Llama-3.2-3BLoss Rate=0.00%2025.12 | 88.62 | |
| Llama-3.2-1BLoss Rate=63.67%2025.12 | 85.34 | |
| Qwen3-0.6BLoss Rate=6.40%2025.12 | 78.67 | |
| DeepSeek-R1-Distill-Qwen-7BLoss Rate=17.59%2025.12 | 14.88 |