new celtic models

This commit is contained in:
Joerg Tiedemann 2020-04-11 15:05:07 +03:00
parent d49a791cc7
commit afa2194c27
2 changed files with 31 additions and 0 deletions

View File

@ -30,3 +30,19 @@
|-----------------------|-------|-------|
| Tatoeba.en.ga | 21.8 | 0.383 |
# opus+techiaith+bt-2020-04-11.zip
* dataset: opus+techiaith+bt
* model: transformer-align
* pre-processing: normalization + SentencePiece
* a sentence initial language token is required in the form of `>>id<<` (id = valid target language ID)
* download: [opus+techiaith+bt-2020-04-11.zip](https://object.pouta.csc.fi/OPUS-MT-models/en-ga+cy+br+gd+kw+gv/opus+techiaith+bt-2020-04-11.zip)
* test set translations: [opus+techiaith+bt-2020-04-11.test.txt](https://object.pouta.csc.fi/OPUS-MT-models/en-ga+cy+br+gd+kw+gv/opus+techiaith+bt-2020-04-11.test.txt)
* test set scores: [opus+techiaith+bt-2020-04-11.eval.txt](https://object.pouta.csc.fi/OPUS-MT-models/en-ga+cy+br+gd+kw+gv/opus+techiaith+bt-2020-04-11.eval.txt)
## Benchmarks
| testset | BLEU | chr-F |
|-----------------------|-------|-------|
| Tatoeba.en.ga | 22.1 | 0.385 |

View File

@ -13,3 +13,18 @@
|-----------------------|-------|-------|
| Tatoeba.ga.en | 25.3 | 0.400 |
# opus+techiaith+bt-2020-04-11.zip
* dataset: opus+techiaith+bt
* model: transformer-align
* pre-processing: normalization + SentencePiece
* download: [opus+techiaith+bt-2020-04-11.zip](https://object.pouta.csc.fi/OPUS-MT-models/ga+cy+br+gd+kw+gv-en/opus+techiaith+bt-2020-04-11.zip)
* test set translations: [opus+techiaith+bt-2020-04-11.test.txt](https://object.pouta.csc.fi/OPUS-MT-models/ga+cy+br+gd+kw+gv-en/opus+techiaith+bt-2020-04-11.test.txt)
* test set scores: [opus+techiaith+bt-2020-04-11.eval.txt](https://object.pouta.csc.fi/OPUS-MT-models/ga+cy+br+gd+kw+gv-en/opus+techiaith+bt-2020-04-11.eval.txt)
## Benchmarks
| testset | BLEU | chr-F |
|-----------------------|-------|-------|
| Tatoeba.ga.en | 25.9 | 0.417 |