Language Generation on P3B3
95.9General ScoreAMALIA-9B-DPO
Evaluation results
| Method | Links | |
|---|---|---|
| AMALIA-9B-DPOModel Type=Fully open models, Training Strategy=DPO2026.03 | 95.9 | |
| AMALIA-9B-SFTModel Type=Fully open models, Training Strategy=SFT2026.03 | 91.3 | |
| Gemma 3-12BModel Type=Open weight models2026.03 | 88.3 | |
| Gemma 2-9BModel Type=Open weight models2026.03 | 72.1 | |
| EuroLLM-9BModel Type=Fully open models2026.03 | 70.5 | |
| Salamandra-7BModel Type=Fully open models2026.03 | 42.7 | |
| Apertus-8BModel Type=Fully open models2026.03 | 28.1 | |
| Llama 3.1-8BModel Type=Open weight models2026.03 | 27.8 | |
| Gervasio-8BModel Type=Open weight models2026.03 | 24.7 | |
| Ministral-8BModel Type=Open weight models2026.03 | 22.1 | |
| Qwen 2.5-7BModel Type=Open weight models2026.03 | 20 | |
| Mistral-7BModel Type=Open weight models2026.03 | 19.2 | |
| Qwen 3-8BModel Type=Open weight models2026.03 | 18.9 | |
| OLMo 2-7BModel Type=Fully open models2026.03 | 18.6 |