Machine Translation On Wmt2015 English German
Metriken
BLEU score
Ergebnisse
Leistungsergebnisse verschiedener Modelle zu diesem Benchmark
Modellname | BLEU score | Paper Title | Repository |
|---|---|---|---|
| Unsupervised attentional encoder-decoder + BPE | 6.89 | Unsupervised Neural Machine Translation | |
| ByteNet | 26.3 | Neural Machine Translation in Linear Time | |
| S2Tree+5gram NPLM | 24.1 | - | - |
| Enc-Dec Att (char) | 23.5 | A Character-Level Decoder without Explicit Segmentation for Neural Machine Translation | |
| BPE word segmentation | 22.8 | Neural Machine Translation of Rare Words with Subword Units | |
| Enc-Dec Att (BPE) | 21.7 | A Character-Level Decoder without Explicit Segmentation for Neural Machine Translation |
0 of 6 row(s) selected.