Machine Translation on FLORES-200 khm_Khmr-English (test)
87.1COMET-22SCALE-refine
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| SCALE-refineLLM=GPT-4, STM=NLLB-3.3B2023.09 | 87.1 | 85.9 | 73.9 | 34.7 | |
| NLLBModel Version=3.3B, Evaluation Protocol=Supervised2023.09 | 86.1 | 85.4 | 72.2 | 35.4 | |
| GPT-4Evaluation Protocol=few-shot2023.09 | 84.6 | 84 | 69.9 | 29.1 | |
| MicrosoftType=Commercial System2023.09 | 80.2 | 80.5 | 63.3 | 25.6 | |
| GPT-3.5Evaluation Protocol=few-shot2023.09 | 73.3 | 73 | 53.2 | 13.9 | |
| M2M100Type=Specialized Model2023.09 | 69.6 | 71.6 | 54 | 17.6 | |
| XGLMEvaluation Protocol=few-shot2023.09 | 48.6 | 53.7 | 21.6 | 0.7 |