translation opus-mt-tc

opus-mt-tc-base-bat-zle

Neural machine translation model for translating from Baltic languages (bat) 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

Usage

A short example code:

from transformers import MarianMTModel, MarianTokenizer

src_text = [
    ">>rus<< Āfrika ir cilvēces šūpulis.",
    ">>ukr<< Tomas yra mūsų kapitonas."
]

model_name = "pytorch-models/opus-mt-tc-base-bat-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:
#     Африка - это колыбель человечества.
#     Томас - наш капітан.

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-base-bat-zle")
print(pipe(">>rus<< Āfrika ir cilvēces šūpulis."))

# expected output: Африка - это колыбель человечества.

Benchmarks

langpair testset chr-F BLEU #sent #words
lav-rus tatoeba-test-v2021-08-07 0.75918 60.5 274 1541
lit-rus tatoeba-test-v2021-08-07 0.72796 54.9 3598 21908
lav-rus flores101-devtest 0.49210 21.1 1012 23295
lav-ukr flores101-devtest 0.48185 19.2 1012 22810
lit-rus flores101-devtest 0.49850 21.3 1012 23295
lit-ukr flores101-devtest 0.49114 19.5 1012 22810

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