opus-mt-eng-nor / README.md
jkorsvik's picture
initial commit
9858ea0
---
language:
- en
- nb
- nn
tags:
- translation
license: apache-2.0
---
### eng-nor
* source language name: English
* target language name: Norwegian
* OPUS readme: [README.md](https://object.pouta.csc.fi/Tatoeba-MT-models/eng-nor/README.md)
* model: transformer-align
* source language code: en
* target language codes: nb, nn
* dataset: opus with backtranslations
* release date: 2021-04-20
* pre-processing: normalization + SentencePiece (spm32k,spm32k)
* download original weights: [opus+bt-2021-04-20.zip](https://object.pouta.csc.fi/Tatoeba-MT-models/eng-nor/opus+bt-2021-04-20.zip/eng-nor/opus+bt-2021-04-20.zip)
* a sentence-initial language token is required in the form of >>id<<(id = valid, usually three-letter target language ID)
* Training data:
* eng-nno: Tatoeba-train (1661769) wikipedia.aa.nno-eng (995603) wikipedia.ab.nno-eng (605107) wikiquote.aa.nno-eng (22626)
* eng-nob: Tatoeba-train (11525999) wikibooks.aa.nob-eng (37901) wikinews.aa.nob-eng (8706) wikipedia.aa.nob-eng (992563) wikipedia.ab.nob-eng (992772) wikipedia.ac.nob-eng (992621) wikipedia.ad.nob-eng (992828) wikipedia.ae.nob-eng (992812) wikipedia.af.nob-eng (976715) wikiquote.aa.nob-eng (10443) wikisource.aa.nob-eng (279891)
* Validation data:
* eng-nno: Tatoeba-dev, 505
* eng-nob: Tatoeba-dev, 5189
* total-size-shuffled: 1505
* devset-selected: top 1505 lines of Tatoeba-dev.src.shuffled
* Test data:
* Tatoeba-test.eng-nno: 460/3428
* Tatoeba-test.eng-nob: 4539/36110
* Tatoeba-test.eng-nor: 4999/39547
* test set translations file: [test.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/eng-nor/opus+bt-2021-04-20.zip/eng-nor/opus+bt-2021-04-20.test.txt)
* test set scores file: [eval.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/eng-nor/opus+bt-2021-04-20.zip/eng-nor/opus+bt-2021-04-20.eval.txt)
* BLEU-scores
|Test set|score|
|---|---|
|Tatoeba-test.eng-nob|56.4|
|Tatoeba-test.eng-nor|55.4|
|Tatoeba-test.eng-nno|40.3|
* chr-F-scores
|Test set|score|
|---|---|
|Tatoeba-test.eng-nob|0.716|
|Tatoeba-test.eng-nor|0.71|
|Tatoeba-test.eng-nno|0.615|
### System Info:
* hf_name: eng-nor
* source_languages: en
* target_languages: nb,nn
* opus_readme_url: https://object.pouta.csc.fi/Tatoeba-MT-models/eng-nor/opus+bt-2021-04-20.zip/README.md
* original_repo: Tatoeba-Challenge
* tags: ['translation']
* languages: ['en', 'nb', 'nn']
* src_constituents: ['eng']
* tgt_constituents: ['nob', 'nno']
* src_multilingual: False
* tgt_multilingual: True
* helsinki_git_sha: 59400fea592520766f9910390155681bc930dbc4
* transformers_git_sha: fd5cdaeea6eafac32e9d967327bfa3dc0e0d962d
* port_machine: DESKTOP-6CPR2HH
* port_time: 2023-01-23-21:07