opus-mt-tc-big-de-zle

Neural machine translation model for translating from German (de) to East Slavic languages (zle).

This model is part of the OPUS-MT project, an effort to make neural machine translation models widely available and accessible for many languages in the world. All models are originally trained using the amazing framework of Marian NMT, an efficient NMT implementation written in pure C++. The models have been converted to pyTorch using the transformers library by huggingface. Training data is taken from OPUS and training pipelines use the procedures of OPUS-MT-train.

@inproceedings{tiedemann-thottingal-2020-opus,
    title = "{OPUS}-{MT} {--} Building open translation services for the World",
    author = {Tiedemann, J{\"o}rg  and Thottingal, Santhosh},
    booktitle = "Proceedings of the 22nd Annual Conference of the European Association for Machine Translation",
    month = nov,
    year = "2020",
    address = "Lisboa, Portugal",
    publisher = "European Association for Machine Translation",
    url = "https://aclanthology.org/2020.eamt-1.61",
    pages = "479--480",
}

@inproceedings{tiedemann-2020-tatoeba,
    title = "The Tatoeba Translation Challenge {--} Realistic Data Sets for Low Resource and Multilingual {MT}",
    author = {Tiedemann, J{\"o}rg},
    booktitle = "Proceedings of the Fifth Conference on Machine Translation",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.wmt-1.139",
    pages = "1174--1182",
}

Model info

This is a multilingual translation model with multiple target languages. A sentence initial language token is required in the form of >>id<< (id = valid target language ID), e.g. >>bel<<

Usage

A short example code:

from transformers import MarianMTModel, MarianTokenizer

src_text = [
    ">>ukr<< Der Soldat hat mir Wasser gegeben.",
    ">>ukr<< Ich will hier nicht essen."
]

model_name = "pytorch-models/opus-mt-tc-big-de-zle"
tokenizer = MarianTokenizer.from_pretrained(model_name)
model = MarianMTModel.from_pretrained(model_name)
translated = model.generate(**tokenizer(src_text, return_tensors="pt", padding=True))

for t in translated:
    print( tokenizer.decode(t, skip_special_tokens=True) )

# expected output:
#     Π‘ΠΎΠ»Π΄Π°Ρ‚ Π΄Π°Π² ΠΌΠ΅Π½Ρ– Π²ΠΎΠ΄Ρƒ.
#     Π― Π½Π΅ Ρ…ΠΎΡ‡Ρƒ Ρ‚ΡƒΡ‚ їсти.

You can also use OPUS-MT models with the transformers pipelines, for example:

from transformers import pipeline
pipe = pipeline("translation", model="Helsinki-NLP/opus-mt-tc-big-de-zle")
print(pipe(">>ukr<< Der Soldat hat mir Wasser gegeben."))

# expected output: Π‘ΠΎΠ»Π΄Π°Ρ‚ Π΄Π°Π² ΠΌΠ΅Π½Ρ– Π²ΠΎΠ΄Ρƒ.

Benchmarks

langpair testset chr-F BLEU #sent #words
deu-bel tatoeba-test-v2021-08-07 0.53128 29.5 551 3601
deu-rus tatoeba-test-v2021-08-07 0.67143 46.1 12800 87296
deu-ukr tatoeba-test-v2021-08-07 0.62737 40.7 10319 56287
deu-rus flores101-devtest 0.54152 26.3 1012 23295
deu-ukr flores101-devtest 0.53286 24.2 1012 22810
deu-rus newstest2012 0.49409 20.8 3003 64790
deu-rus newstest2013 0.52631 24.9 3000 58560

Acknowledgements

The work is supported by the European Language Grid as pilot project 2866, by the FoTran project, funded by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 771113), and the MeMAD project, funded by the European Union’s Horizon 2020 Research and Innovation Programme under grant agreement No 780069. We are also grateful for the generous computational resources and IT infrastructure provided by CSC -- IT Center for Science, Finland.

Model conversion info

  • transformers version: 4.16.2
  • OPUS-MT git hash: 1bdabf7
  • port time: Thu Mar 24 01:29:09 EET 2022
  • port machine: LM0-400-22516.local
Downloads last month
43
Safetensors
Model size
239M params
Tensor type
FP16
Β·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Spaces using Helsinki-NLP/opus-mt-tc-big-de-zle 7

Evaluation results