Named Entity Recognition on UFHealth Opioid use dataset
91.8F1 ScoreFull FT
Evaluation results
| Method | Links | |
|---|---|---|
| Full FTBase model=gpt-oss 20B, Trainable Parameters=100.0%2026.04 | 91.8 | |
| MPTBase model=gpt-oss 20B, Trainable Parameters=<0.05%2026.04 | 91.4 | |
| Full FTBase model=Meditron3 8B, Trainable Parameters=100.0%2026.04 | 90.4 | |
| MPTBase model=Meditron3 8B, Trainable Parameters=<0.05%2026.04 | 89.5 | |
| LoRABase model=gpt-oss 20B, Trainable Parameters=~2.50%2026.04 | 89.5 | |
| LoRABase model=Meditron3 8B, Trainable Parameters=~2.50%2026.04 | 87.9 | |
| Full FTBase model=LLaMA 3.1 8B, Trainable Parameters=100.0%2026.04 | 87.7 | |
| MPTBase model=LLaMA 3.1 8B, Trainable Parameters=<0.05%2026.04 | 86.8 | |
| LoRABase model=LLaMA 3.1 8B, Trainable Parameters=~2.50%2026.04 | 85 | |
| PTBase model=gpt-oss 20B, Trainable Parameters=<0.05%2026.04 | 84.8 | |
| PTBase model=Meditron3 8B, Trainable Parameters=<0.05%2026.04 | 83 | |
| PTBase model=LLaMA 3.1 8B, Trainable Parameters=<0.05%2026.04 | 79.9 |