Back to all models
translation mask_token:
Query this model
🔥 This model is currently loaded and running on the Inference API. ⚠️ This model could not be loaded by the inference API. ⚠️ This model can be loaded on the Inference API on-demand.
JSON Output
API endpoint  

⚡️ Upgrade your account to access the Inference API

Share Copied link to clipboard

Monthly model downloads

Helsinki-NLP/opus-mt-en-zle Helsinki-NLP/opus-mt-en-zle
N/a downloads
last 30 days

pytorch

tf

Contributed by

Language Technology Research Group at the University of Helsinki university
1 team member · 1325 models

How to use this model directly from the 🤗/transformers library:

			
Copy to clipboard
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM tokenizer = AutoTokenizer.from_pretrained("Helsinki-NLP/opus-mt-en-zle") model = AutoModelForSeq2SeqLM.from_pretrained("Helsinki-NLP/opus-mt-en-zle")
Uploaded in S3

eng-zle

  • source group: English

  • target group: East Slavic languages

  • OPUS readme: eng-zle

  • model: transformer

  • source language(s): eng

  • target language(s): bel bel_Latn orv_Cyrl rue rus ukr

  • model: transformer

  • pre-processing: normalization + SentencePiece (spm32k,spm32k)

  • a sentence initial language token is required in the form of >>id<< (id = valid target language ID)

  • download original weights: opus2m-2020-08-02.zip

  • test set translations: opus2m-2020-08-02.test.txt

  • test set scores: opus2m-2020-08-02.eval.txt

Benchmarks

testset BLEU chr-F
newstest2012-engrus.eng.rus 27.4 0.550
newstest2013-engrus.eng.rus 21.4 0.493
newstest2015-enru-engrus.eng.rus 24.2 0.534
newstest2016-enru-engrus.eng.rus 23.3 0.518
newstest2017-enru-engrus.eng.rus 25.3 0.541
newstest2018-enru-engrus.eng.rus 22.4 0.527
newstest2019-enru-engrus.eng.rus 24.1 0.505
Tatoeba-test.eng-bel.eng.bel 20.8 0.471
Tatoeba-test.eng.multi 37.2 0.580
Tatoeba-test.eng-orv.eng.orv 0.6 0.130
Tatoeba-test.eng-rue.eng.rue 1.4 0.168
Tatoeba-test.eng-rus.eng.rus 41.3 0.616
Tatoeba-test.eng-ukr.eng.ukr 38.7 0.596

System Info: