Machine Translation on WMT22 Ru-Sah (test)
54.5COMET ScoreDUAL-REFLECT
Evaluation results
| Method | Links | ||
|---|---|---|---|
| DUAL-REFLECTInference Methodology=DUAL-REFLECT2024.06 | 54.5 | 76.9 | |
| ChatGPT + MADInference Methodology=MAD2024.06 | 53.5 | 76.4 | |
| ChatGPT + 5-shotPrompting Strategy=Few-shot, Shots=52024.06 | 53.1 | 75.4 | |
| ChatGPTPrompting Strategy=Zero-shot2024.06 | 52.8 | 73.2 |