Machine Translation on Arabic-Portuguese
29.5BLEUPIVOTE
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| PIVOTEMerging Module=GPT-4o, Candidates=Direct & English Pivot2025.02 | 29.5 | 53.16 | 86.03 | |
| PIVOTEMerging Module=GPT-4o, Candidates=top32025.02 | 28.62 | 52.53 | 85.87 | |
| PIVOTEMerging Module=GPT-4, Candidates=Direct & English Pivot2025.02 | 27.98 | 52.41 | 85.27 | |
| GPT-4otype=Standalone NMT System2025.02 | 27.28 | 52.57 | 85.9 | |
| NLLBtype=Standalone NMT System2025.02 | 27.25 | 50.35 | 84.21 | |
| PIVOTEMerging Module=GPT-4, Candidates=top32025.02 | 27.22 | 51.73 | 85.65 | |
| GPT-4type=Standalone NMT System2025.02 | 25.82 | 51.89 | 85.46 | |
| MBRtype=Prior Ensemble Method2025.02 | 25.45 | 51.78 | 85.55 | |
| PIVOTEMerging Module=Llama-3, Candidates=top32025.02 | 23.41 | 45.95 | 81.66 | |
| PIVOTEMerging Module=Llama-3, Candidates=Direct & English Pivot2025.02 | 21.35 | 43.75 | 79.71 | |
| Llama-3type=Standalone NMT System2025.02 | 18.78 | 40.2 | 78.73 | |
| Vicunatype=Standalone NMT System2025.02 | 17.64 | 38.44 | 76.01 | |
| Baizetype=Standalone NMT System2025.02 | 16.56 | 36.67 | 76.87 | |
| LLM-Blendertype=Prior Ensemble Method2025.02 | 11.8 | 29.85 | 67.95 | |
| EVAtype=Prior Ensemble Method2025.02 | 9.77 | 28.4 | 68.75 |