Machine Translation on IWSLT En-Fr 2017 (test)
40.02BLEUBPE-dropout
Evaluation results
| Method | Links | |
|---|---|---|
| BPE-dropoutSegmentation=BPE-dropout2019.10 | 40.02 | |
| Kudo (2018)Segmentation=Unigram Language Model2019.10 | 39.45 | |
| BPESegmentation=BPE2019.10 | 39.37 | |
| FAASTBackbone=Qwen2.5-7B-Instruct, Protocol=FAAST, Data Setting=full2026.05 | 37.08 | |
| FAASTBackbone=Qwen2.5-7B-Instruct, Protocol=FAAST, Data Setting=1-shot2026.05 | 35.67 | |
| In-Context LearningBackbone=Qwen2.5-7B-Instruct, Protocol=In-Context Learning, Data Setting=1-shot2026.05 | 35.45 | |
| FAASTBackbone=Qwen2.5-3B-Instruct, Protocol=FAAST, Data Setting=full2026.05 | 35.09 | |
| Qwen2.5-7B-InstructBackbone=Qwen2.5-7B-Instruct, Protocol=zero-shot, Data Setting=full2026.05 | 34.82 | |
| In-Context LearningBackbone=Qwen2.5-3B-Instruct, Protocol=In-Context Learning, Data Setting=1-shot2026.05 | 31.85 | |
| FAASTBackbone=Qwen2.5-3B-Instruct, Protocol=FAAST, Data Setting=1-shot2026.05 | 31.12 | |
| Qwen2.5-3B-InstructBackbone=Qwen2.5-3B-Instruct, Protocol=zero-shot, Data Setting=full2026.05 | 30.56 |