Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,36 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: cc-by-nc-sa-4.0
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-nc-sa-4.0
|
| 3 |
+
language:
|
| 4 |
+
- cs
|
| 5 |
+
- en
|
| 6 |
+
pipeline_tag: translation
|
| 7 |
+
---
|
| 8 |
+
|
| 9 |
+
# Opus Tatoeba | Czech -> English
|
| 10 |
+
|
| 11 |
+
* dataset: opus
|
| 12 |
+
* model: transformer
|
| 13 |
+
* source language(s): ces
|
| 14 |
+
* target language(s): eng
|
| 15 |
+
* model: transformer
|
| 16 |
+
* pre-processing: normalization + SentencePiece (spm32k,spm32k)
|
| 17 |
+
* download: [opus-2021-02-19.zip](https://object.pouta.csc.fi/Tatoeba-MT-models/ces-eng/opus-2021-02-19.zip)
|
| 18 |
+
* test set translations: [opus-2021-02-19.test.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/ces-eng/opus-2021-02-19.test.txt)
|
| 19 |
+
* test set scores: [opus-2021-02-19.eval.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/ces-eng/opus-2021-02-19.eval.txt)
|
| 20 |
+
|
| 21 |
+
## Benchmarks
|
| 22 |
+
|
| 23 |
+
| testset | BLEU | chr-F | #sent | #words | BP |
|
| 24 |
+
|---------|-------|-------|-------|--------|----|
|
| 25 |
+
| newssyscomb2009.ces-eng | 27.7 | 0.551 | 502 | 11821 | 0.971 |
|
| 26 |
+
| newstest2009.ces-eng | 27.2 | 0.550 | 2525 | 65402 | 0.970 |
|
| 27 |
+
| newstest2010.ces-eng | 27.3 | 0.559 | 2489 | 61724 | 0.978 |
|
| 28 |
+
| newstest2011.ces-eng | 28.0 | 0.557 | 3003 | 74681 | 0.990 |
|
| 29 |
+
| newstest2012.ces-eng | 27.2 | 0.552 | 3003 | 72812 | 1.000 |
|
| 30 |
+
| newstest2013.ces-eng | 30.7 | 0.572 | 3000 | 64505 | 1.000 |
|
| 31 |
+
| newstest2014-csen.ces-eng | 34.2 | 0.614 | 3003 | 68065 | 0.999 |
|
| 32 |
+
| newstest2015-encs.ces-eng | 30.7 | 0.568 | 2656 | 53572 | 0.975 |
|
| 33 |
+
| newstest2016-encs.ces-eng | 32.4 | 0.589 | 2999 | 64670 | 0.998 |
|
| 34 |
+
| newstest2017-encs.ces-eng | 28.9 | 0.559 | 3005 | 61725 | 0.996 |
|
| 35 |
+
| newstest2018-encs.ces-eng | 30.4 | 0.568 | 2983 | 63496 | 0.991 |
|
| 36 |
+
| Tatoeba-test.ces-eng | 56.9 | 0.719 | 10000 | 75376 | 0.962 |
|