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: 2024-05-29
License: Apache-2.0
Language(s):
- Source Language(s): deu eng fra por spa
- Target Language(s): bru cmo hoc jun kha khm kxm mnw ngt sat vie wbm
- Valid Target Language Labels: >>aem<< >>alk<< >>aml<< >>asr<< >>bbh<< >>bdq<< >>bfw<< >>bgk<< >>bgl<< >>bix<< >>biy<< >>blr<< >>brb<< >>bru<< >>brv<< >>btq<< >>caq<< >>cbn<< >>cdz<< >>cma<< >>cmo<< >>cog<< >>crv<< >>crw<< >>cua<< >>cwg<< >>dnu<< >>ekl<< >>gaq<< >>gbj<< >>hal<< >>hld<< >>hnu<< >>hoc<< >>hoc_Wara<< >>hre<< >>huo<< >>irr<< >>jah<< >>jeh<< >>jhi<< >>jun<< >>juy<< >>kdt<< >>kfp<< >>kfq<< >>kha<< >>khf<< >>khm<< >>khr<< >>kjg<< >>kjm<< >>knq<< >>kns<< >>kpm<< >>krr<< >>krv<< >>ksz<< >>kta<< >>ktv<< >>kuf<< >>kxm<< >>kxy<< >>lbn<< >>lbo<< >>lcp<< >>lnh<< >>lwl<< >>lyg<< >>mef<< >>mhe<< >>mjx<< >>mlf<< >>mmj<< >>mml<< >>mng<< >>mnn<< >>mnq<< >>mnw<< >>moo<< >>mqt<< >>mra<< >>mtq<< >>mzt<< >>ncb<< >>ncq<< >>nev<< >>ngt<< >>ngt_Latn<< >>nik<< >>nuo<< >>nyl<< >>omx<< >>oog<< >>oyb<< >>pac<< >>pbv<< >>pcb<< >>pce<< >>pcj<< >>phg<< >>pkt<< >>pll<< >>ply<< >>pnx<< >>prk<< >>prt<< >>puo<< >>rbb<< >>ren<< >>ril<< >>rka<< >>rmx<< >>sat<< >>sat_Latn<< >>sbo<< >>scb<< >>scq<< >>sct<< >>sea<< >>sed<< >>sii<< >>smu<< >>spu<< >>sqq<< >>srb<< >>ssm<< >>sss<< >>stg<< >>sti<< >>stt<< >>stu<< >>syo<< >>sza<< >>szc<< >>tdf<< >>tdr<< >>tea<< >>tef<< >>thm<< >>tkz<< >>tlq<< >>tmo<< >>tnz<< >>tou<< >>tpu<< >>trd<< >>tth<< >>tto<< >>tyh<< >>unr<< >>uuu<< >>vie<< >>vwa<< >>wbm<< >>xao<< >>xkk<< >>xnh<< >>xxx<< >>yin<< >>zng<<
Original Model: opusTCv20230926max50+bt+jhubc_transformer-big_2024-05-29.zip
Resources for more information:

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. >>bru<<

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 = [
    ">>khm<< Der Junge wirft einen Stein.",
    ">>vie<< ¿Y tú?"
]

model_name = "pytorch-models/opus-mt-tc-bible-big-deu_eng_fra_por_spa-aav"
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:
#     ក្មេងប្រុស នោះ យក ដុំ ថ្ម គប់ ។
#     Còn anh thì sao?

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-bible-big-deu_eng_fra_por_spa-aav")
print(pipe(">>khm<< Der Junge wirft einen Stein."))

# expected output: ក្មេងប្រុស នោះ យក ដុំ ថ្ម គប់ ។

Training

Data: opusTCv20230926max50+bt+jhubc (source)
Pre-processing: SentencePiece (spm32k,spm32k)
Model Type: transformer-big
Original MarianNMT Model: opusTCv20230926max50+bt+jhubc_transformer-big_2024-05-29.zip
Training Scripts: GitHub Repo

Evaluation

Model scores at the OPUS-MT dashboard
test set translations: opusTCv20230926max50+bt+jhubc_transformer-big_2024-05-29.test.txt
test set scores: opusTCv20230926max50+bt+jhubc_transformer-big_2024-05-29.eval.txt
benchmark results: benchmark_results.txt
benchmark output: benchmark_translations.zip

langpair	testset	chr-F	BLEU	#sent	#words
deu-vie	tatoeba-test-v2021-08-07	0.45795	25.6	400	3768
eng-hoc	tatoeba-test-v2021-08-07	6.438	0.2	660	2591
eng-kha	tatoeba-test-v2021-08-07	5.741	0.0	1314	9269
eng-vie	tatoeba-test-v2021-08-07	0.56461	39.4	2500	24427
fra-vie	tatoeba-test-v2021-08-07	0.52806	35.2	1299	13219
spa-vie	tatoeba-test-v2021-08-07	0.52131	34.2	594	4740
deu-vie	flores101-devtest	0.53381	33.8	1012	33331
eng-khm	flores101-devtest	0.42302	1.3	1012	7006
eng-vie	flores101-devtest	0.59621	42.1	1012	33331
fra-khm	flores101-devtest	0.40042	2.2	1012	7006
por-khm	flores101-devtest	0.40585	2.1	1012	7006
por-vie	flores101-devtest	0.54919	36.0	1012	33331
spa-vie	flores101-devtest	0.49921	27.8	1012	33331
deu-vie	flores200-devtest	0.53671	34.0	1012	33331
eng-khm	flores200-devtest	0.42148	1.3	1012	7006
eng-vie	flores200-devtest	0.59842	42.4	1012	33331
fra-vie	flores200-devtest	0.54101	34.6	1012	33331
por-khm	flores200-devtest	0.40832	1.9	1012	7006
por-vie	flores200-devtest	0.54970	36.1	1012	33331
spa-vie	flores200-devtest	0.50025	28.1	1012	33331
deu-khm	ntrex128	0.44903	3.5	1997	15866
deu-vie	ntrex128	0.52124	31.4	1997	64655
eng-khm	ntrex128	0.50494	1.6	1997	15866
eng-vie	ntrex128	3.831	0.0	1997	64655
fra-khm	ntrex128	0.43841	2.4	1997	15866
fra-vie	ntrex128	0.52044	31.8	1997	64655
por-khm	ntrex128	0.46655	2.5	1997	15866
por-vie	ntrex128	0.53060	33.3	1997	64655
spa-khm	ntrex128	0.46443	2.7	1997	15866
spa-vie	ntrex128	0.53293	33.4	1997	64655
eng-khm	tico19-test	0.47806	2.5	2100	15810
fra-khm	tico19-test	3.268	1.0	2100	15810
por-khm	tico19-test	3.900	1.1	2100	15810
spa-khm	tico19-test	3.784	1.0	2100	15810

Citation Information

Publications: Democratizing neural machine translation with OPUS-MT and 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.)

bibtex

@article{tiedemann2023democratizing,
  title={Democratizing neural machine translation with {OPUS-MT}},
  author={Tiedemann, J{\"o}rg and Aulamo, Mikko and Bakshandaeva, Daria and Boggia, Michele and Gr{\"o}nroos, Stig-Arne and Nieminen, Tommi and Raganato, Alessandro and Scherrer, Yves and Vazquez, Raul and Virpioja, Sami},
  journal={Language Resources and Evaluation},
  number={58},
  pages={713--755},
  year={2023},
  publisher={Springer Nature},
  issn={1574-0218},
  doi={10.1007/s10579-023-09704-w}
}

@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 HPLT project, funded by the European Union’s Horizon Europe research and innovation programme under grant agreement No 101070350. We are also grateful for the generous computational resources and IT infrastructure provided by CSC -- IT Center for Science, Finland, and the EuroHPC supercomputer LUMI.

Model conversion info

transformers version: 4.45.1
OPUS-MT git hash: 0882077
port time: Tue Oct 8 08:57:20 EEST 2024
port machine: LM0-400-22516.local

Helsinki-NLP/opus-mt-tc-bible-big-deu_eng_fra_por_spa-aav

opus-mt-tc-bible-big-deu_eng_fra_por_spa-aav

Table of Contents

Model Details