Idiomatic Translation on Mixed Dataset en-ta
1.87LLM-eval ScoreIdiomCE
Evaluation results
| Method | Links | ||
|---|---|---|---|
| IdiomCEModel=GPT-4o2025.05 | 1.87 | 0.67 | |
| DirectModel=GPT-4o2025.05 | 1.741 | 0.72 | |
| IdiomCEModel=Gemma2-9b-it2025.05 | 1.63 | 0.67 | |
| DirectModel=Gemma2-9b-it2025.05 | 1.56 | 0.71 | |
| IdiomCEModel=LLama-3.1-8B2025.05 | 1.25 | 0.57 | |
| DirectModel=Indictrans22025.05 | 1.243 | 0.769 | |
| DirectModel=NLLB-2002025.05 | 1.18 | 0.691 | |
| DirectModel=LLama-3.1-8B2025.05 | 1.16 | 0.62 | |
| IdiomCEModel=LLama-3.2-3B2025.05 | 1.105 | 0.51 | |
| DirectModel=LLama-3.2-3B2025.05 | 1.04 | 0.52 |