Language Generation on PT-Exams Open Questions
77.3ScoreQwen 3-8B
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen 3-8BModel Type=Open weight models2026.03 | 77.3 | |
| Gemma 3-12BModel Type=Open weight models2026.03 | 76.6 | |
| Gemma 2-9BModel Type=Open weight models2026.03 | 69.7 | |
| AMALIA-9B-DPOModel Type=Fully open models, Training Strategy=DPO2026.03 | 66 | |
| AMALIA-9B-SFTModel Type=Fully open models, Training Strategy=SFT2026.03 | 62 | |
| Ministral-8BModel Type=Open weight models2026.03 | 62 | |
| Qwen 2.5-7BModel Type=Open weight models2026.03 | 56.8 | |
| EuroLLM-9BModel Type=Fully open models2026.03 | 56.1 | |
| Apertus-8BModel Type=Fully open models2026.03 | 54.7 | |
| Llama 3.1-8BModel Type=Open weight models2026.03 | 53.8 | |
| Gervasio-8BModel Type=Open weight models2026.03 | 53.2 | |
| Mistral-7BModel Type=Open weight models2026.03 | 44.6 | |
| OLMo 2-7BModel Type=Fully open models2026.03 | 43 | |
| Salamandra-7BModel Type=Fully open models2026.03 | 34.1 |