Machine Translation (Zh-En) on WMT Original 2017-2019 (test)
26.6BLEUTST BT
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| TST BTTraining Data=WMT 2017 Bitext + TST BT2023.06 | 26.6 | 54.9 | 47.8 | 69.5 | |
| Beam BTTraining Data=WMT 2017 Bitext + Beam BT2023.06 | 26.4 | 55.1 | 46.4 | 69.1 | |
| BitextTraining Data=WMT 2017 Bitext2023.06 | 24.7 | 53.4 | 43.5 | 68 | |
| BitextTraining Data=WMT 2017 Bitext2023.06 | 0.238 | 0.53 | 0.342 | 0.659 | |
| Beam BTTraining Data=WMT 2017 Bitext + Beam BT2023.06 | 0.237 | 0.535 | 0.362 | 0.666 | |
| TST BTTraining Data=WMT 2017 Bitext + TST BT2023.06 | 0.235 | 0.531 | 0.374 | 0.67 |