mirror of
https://github.com/Helsinki-NLP/OPUS-MT-train.git
synced 2024-09-11 20:27:19 +03:00
| .. | ||
| README.md | ||
opus-2019-12-04.zip
- dataset: opus
- model: transformer
- pre-processing: normalization + tokenization + BPE
- download: opus-2019-12-04.zip
- test set translations: opus-2019-12-04.test.txt
- test set scores: opus-2019-12-04.eval.txt
Benchmarks
| testset | BLEU | chr-F |
|---|---|---|
| Tatoeba.en.eo | 32.4 | 0.533 |
opus-2019-12-18.zip
- dataset: opus
- model: transformer-align
- pre-processing: normalization + SentencePiece
- download: opus-2019-12-18.zip
- test set translations: opus-2019-12-18.test.txt
- test set scores: opus-2019-12-18.eval.txt
Benchmarks
| testset | BLEU | chr-F |
|---|---|---|
| Tatoeba.en.eo | 49.5 | 0.682 |
opus-2020-05-23.zip
- dataset: opus
- model: transformer-align
- source language(s): en
- target language(s): eo
- model: transformer-align
- pre-processing: normalization + SentencePiece (spm4k,spm4k)
- download: opus-2020-05-23.zip
- test set translations: opus-2020-05-23.test.txt
- test set scores: opus-2020-05-23.eval.txt
Training data: opus+bt
- en-eo: bible-uedin (61412) Books (1492) GlobalVoices (2779) GNOME (19165) KDE4 (37316) OpenSubtitles (51388) QED (15746) Ubuntu (4441)
- en-eo: total size = 193739
- unused dev/test data is added to training data
- total size (opus+bt): 405729
Validation data
-
en-eo: Tatoeba
-
devset = top 5000 lines of Tatoeba.src.shuffled!
-
testset = next 5000 lines of Tatoeba.src.shuffled!
-
remaining lines are added to traindata
Benchmarks
| testset | BLEU | chr-F |
|---|---|---|
| Tatoeba.en.eo | 50.7 | 0.693 |