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. Model Description:

Developed by: Language Technology Research Group at the University of Helsinki
Model Type: Translation (transformer-big)
Release: 2022-07-29
License: CC-BY-4.0
Language(s):
- Source Language(s): dan fao isl nno nob nor swe
- Target Language(s): dan isl nno nob nor swe
- Language Pair(s): dan-isl dan-nob dan-swe isl-dan isl-nob isl-swe nno-nob nob-dan nob-isl nob-nno nob-swe swe-dan swe-isl swe-nob
- Valid Target Language Labels: >>dan<< >>fao<< >>isl<< >>jut<< >>nno<< >>nob<< >>non<< >>nrn<< >>ovd<< >>qer<< >>rmg<< >>swe<<
Original Model: opusTCv20210807_transformer-big_2022-07-29.zip
Resources for more information:
- OPUS-MT-train GitHub Repo
- More information about released models for this language pair: OPUS-MT gmq-gmq README
- More information about MarianNMT models in the transformers library
- [Tatoeba Translation Challenge](https://github.com/Helsinki-NLP/Tatoeba-Challenge/

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. >>dan<<

Uses

This model can be used for translation and text-to-text generation.

Risks, Limitations and Biases

CONTENT WARNING: Readers should be aware that the model is trained on various public data sets that may contain content that is disturbing, offensive, and can propagate historical and current stereotypes.

Significant research has explored bias and fairness issues with language models (see, e.g., Sheng et al. (2021) and Bender et al. (2021)).

How to Get Started With the Model

A short example code:

python

from transformers import MarianMTModel, MarianTokenizer

src_text = [
    ">>fao<< Jeg er bange for kakerlakker.",
    ">>nob<< Vladivostok är en stad i Ryssland."
]

model_name = "pytorch-models/opus-mt-tc-big-gmq-gmq"
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:
#     Tað eru uml.
#     Vladivostok er en by i Russland.

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

python

from transformers import pipeline
pipe = pipeline("translation", model="Helsinki-NLP/opus-mt-tc-big-gmq-gmq")
print(pipe(">>fao<< Jeg er bange for kakerlakker."))

# expected output: Tað eru uml.

Training

Data: opusTCv20210807 (source)
Pre-processing: SentencePiece (spm32k,spm32k)
Model Type: transformer-big
Original MarianNMT Model: opusTCv20210807_transformer-big_2022-07-29.zip
Training Scripts: GitHub Repo

Evaluation

test set translations: opusTCv20210807_transformer-big_2022-07-29.test.txt
test set scores: opusTCv20210807_transformer-big_2022-07-29.eval.txt
benchmark results: benchmark_results.txt
benchmark output: benchmark_translations.zip

langpair	testset	chr-F	BLEU	#sent	#words
dan-nob	tatoeba-test-v2021-08-07	0.87556	78.2	1299	9620
dan-swe	tatoeba-test-v2021-08-07	0.83556	72.5	1549	10060
nno-nob	tatoeba-test-v2021-08-07	0.88349	78.9	467	3129
nob-dan	tatoeba-test-v2021-08-07	0.85345	73.9	1299	9794
nob-nno	tatoeba-test-v2021-08-07	0.74571	55.2	466	3141
nob-swe	tatoeba-test-v2021-08-07	0.84747	73.9	563	3698
swe-dan	tatoeba-test-v2021-08-07	0.83392	72.6	1549	10239
swe-nob	tatoeba-test-v2021-08-07	0.85815	76.3	563	3708
isl-swe	europeana2021	0.45562	22.2	563	10293
nob-isl	europeana2021	0.54171	29.7	538	9932
nob-swe	europeana2021	0.73891	54.0	538	9885
dan-isl	flores101-devtest	0.50227	22.2	1012	22834
dan-nob	flores101-devtest	0.58445	28.6	1012	23873
dan-swe	flores101-devtest	0.65000	38.5	1012	23121
isl-dan	flores101-devtest	0.53630	27.2	1012	24638
isl-nob	flores101-devtest	0.49434	20.5	1012	23873
isl-swe	flores101-devtest	0.53373	26.0	1012	23121
nob-dan	flores101-devtest	0.59657	31.7	1012	24638
nob-isl	flores101-devtest	0.47432	18.9	1012	22834
nob-swe	flores101-devtest	0.60030	31.3	1012	23121
swe-dan	flores101-devtest	0.64340	39.0	1012	24638
swe-isl	flores101-devtest	0.49590	21.7	1012	22834
swe-nob	flores101-devtest	0.58336	28.9	1012	23873

Citation Information

Publications: OPUS-MT – Building open translation services for the World and The Tatoeba Translation Challenge – Realistic Data Sets for Low Resource and Multilingual MT (Please, cite if you use this model.)

@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",
}

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: 8b9f0b0
port time: Fri Aug 12 23:59:02 EEST 2022
port machine: LM0-400-22516.local

Helsinki-NLP/opus-mt-tc-big-gmq-gmq

opus-mt-tc-big-gmq-gmq

Table of Contents

Model Details