Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,38 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: cc-by-nc-sa-4.0
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-nc-sa-4.0
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
- de
|
| 6 |
+
pipeline_tag: translation
|
| 7 |
+
---
|
| 8 |
+
|
| 9 |
+
# Opus Tatoeba | English -> German
|
| 10 |
+
|
| 11 |
+
* dataset: opus
|
| 12 |
+
* model: transformer
|
| 13 |
+
* source language(s): eng
|
| 14 |
+
* target language(s): deu
|
| 15 |
+
* model: transformer
|
| 16 |
+
* pre-processing: normalization + SentencePiece (spm32k,spm32k)
|
| 17 |
+
* download: [opus-2021-02-22.zip](https://object.pouta.csc.fi/Tatoeba-MT-models/eng-deu/opus-2021-02-22.zip)
|
| 18 |
+
* test set translations: [opus-2021-02-22.test.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/eng-deu/opus-2021-02-22.test.txt)
|
| 19 |
+
* test set scores: [opus-2021-02-22.eval.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/eng-deu/opus-2021-02-22.eval.txt)
|
| 20 |
+
|
| 21 |
+
## Benchmarks
|
| 22 |
+
|
| 23 |
+
| testset | BLEU | chr-F | #sent | #words | BP |
|
| 24 |
+
|---------|-------|-------|-------|--------|----|
|
| 25 |
+
| newssyscomb2009.eng-deu | 23.3 | 0.539 | 502 | 11271 | 0.990 |
|
| 26 |
+
| news-test2008.eng-deu | 23.9 | 0.533 | 2051 | 47427 | 1.000 |
|
| 27 |
+
| newstest2009.eng-deu | 22.7 | 0.533 | 2525 | 62816 | 0.999 |
|
| 28 |
+
| newstest2010.eng-deu | 25.9 | 0.550 | 2489 | 61511 | 0.966 |
|
| 29 |
+
| newstest2011.eng-deu | 22.9 | 0.528 | 3003 | 72981 | 0.993 |
|
| 30 |
+
| newstest2012.eng-deu | 23.8 | 0.530 | 3003 | 72886 | 0.972 |
|
| 31 |
+
| newstest2013.eng-deu | 27.6 | 0.558 | 3000 | 63737 | 0.983 |
|
| 32 |
+
| newstest2014-deen.eng-deu | 29.6 | 0.595 | 3003 | 62964 | 1.000 |
|
| 33 |
+
| newstest2015-ende.eng-deu | 32.0 | 0.601 | 2169 | 44260 | 1.000 |
|
| 34 |
+
| newstest2016-ende.eng-deu | 37.9 | 0.644 | 2999 | 62670 | 0.992 |
|
| 35 |
+
| newstest2017-ende.eng-deu | 30.6 | 0.593 | 3004 | 61291 | 1.000 |
|
| 36 |
+
| newstest2018-ende.eng-deu | 46.4 | 0.697 | 2998 | 64276 | 0.999 |
|
| 37 |
+
| newstest2019-ende.eng-deu | 42.4 | 0.664 | 1997 | 48969 | 0.990 |
|
| 38 |
+
| Tatoeba-test.eng-deu | 45.8 | 0.655 | 10000 | 83347 | 0.995 |
|