Relation Extraction on UFHealth Opioid Use Dataset
89.9F1 ScoreFull FT
Evaluation results
| Method | Links | |
|---|---|---|
| Full FTBase model=gpt-oss 20B, Trainable Parameters=100.0%2026.04 | 89.9 | |
| MPTBase model=gpt-oss 20B, Trainable Parameters=<0.05%2026.04 | 89.3 | |
| Full FTBase model=Meditron3 8B, Trainable Parameters=100.0%2026.04 | 88.2 | |
| MPTBase model=Meditron3 8B, Trainable Parameters=<0.05%2026.04 | 87.5 | |
| LoRABase model=gpt-oss 20B, Trainable Parameters=~2.50%2026.04 | 87.3 | |
| Full FTBase model=LLaMA 3.1 8B, Trainable Parameters=100.0%2026.04 | 85.8 | |
| LoRABase model=Meditron3 8B, Trainable Parameters=~2.50%2026.04 | 85.4 | |
| MPTBase model=LLaMA 3.1 8B, Trainable Parameters=<0.05%2026.04 | 85 | |
| LoRABase model=LLaMA 3.1 8B, Trainable Parameters=~2.50%2026.04 | 82.8 | |
| PTBase model=gpt-oss 20B, Trainable Parameters=<0.05%2026.04 | 82.2 | |
| PTBase model=Meditron3 8B, Trainable Parameters=<0.05%2026.04 | 80.1 | |
| PTBase model=LLaMA 3.1 8B, Trainable Parameters=<0.05%2026.04 | 77.3 |