Machine Translation on FLORES-200 amh_Ethi-English (test)
88COMET-22SCALE-refine
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| SCALE-refineLLM=GPT-4, STM=NLLB-3.3B2023.09 | 88 | 85.3 | 75.7 | 37.6 | |
| MicrosoftType=Commercial System2023.09 | 87.5 | 84.6 | 74.7 | 41.9 | |
| NLLBModel Version=3.3B, Evaluation Protocol=Supervised2023.09 | 86.9 | 84.5 | 73.6 | 36.4 | |
| GPT-4Evaluation Protocol=few-shot2023.09 | 83.2 | 81.9 | 67.3 | 27.1 | |
| M2M100Type=Specialized Model2023.09 | 72.3 | 72 | 54.8 | 18.5 | |
| GPT-3.5Evaluation Protocol=few-shot2023.09 | 58.8 | 54.2 | 31.7 | 3.4 | |
| XGLMEvaluation Protocol=few-shot2023.09 | 50.2 | 43.9 | 17.8 | 0.1 |