Helsinki-NLP_opus-mt-tc-big-en-bg

---

language:

• bg
• en tags:
• translation
• opus-mt-tc license: cc-by-4.0 model-index:
• name: opus-mt-tc-big-en-bg results:
  • task: name: Translation eng-bul type: translation args: eng-bul dataset: name: flores101-devtest type: flores_101 args: eng bul devtest metrics:
    • name: BLEU type: bleu value: 44.9
  • task: name: Translation eng-bul type: translation args: eng-bul dataset: name: tatoeba-test-v2021-08-07 type: tatoeba_mt args: eng-bul metrics:
    • name: BLEU type: bleu value: 51.5

---

opus-mt-tc-big-en-bg

Neural machine translation model for translating from English (en) to Bulgarian (bg).

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.

• 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",
}

Model info

• Release: 2022-02-25
• source language(s): eng
• target language(s): bul
• model: transformer-big
• data: opusTCv20210807+bt (source)
• tokenization: SentencePiece (spm32k,spm32k)
• original model: opusTCv20210807+bt_transformer-big_2022-02-25.zip
• more information released models: OPUS-MT eng-bul README

Usage

A short example code:

from transformers import MarianMTModel, MarianTokenizer

src_text = [
    "2001 is the year when the 21st century begins.",
    "This is Copacabana!"
]

model_name = "pytorch-models/opus-mt-tc-big-en-bg"
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:
#     2001 е годината, в която започва 21-ви век.
#     Това е Копакабана!

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-en-bg")
print(pipe("2001 is the year when the 21st century begins."))

# expected output: 2001 е годината, в която започва 21-ви век.

Benchmarks

• test set translations: opusTCv20210807+bt_transformer-big_2022-02-25.test.txt
• test set scores: opusTCv20210807+bt_transformer-big_2022-02-25.eval.txt
• benchmark results: benchmark_results.txt
• benchmark output: benchmark_translations.zip

langpair	testset	chr-F	BLEU	#sent	#words
eng-bul	tatoeba-test-v2021-08-07	0.68987	51.5	10000	69504
eng-bul	flores101-devtest	0.69891	44.9	1012	24700

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: 3405783
• port time: Wed Apr 13 16:29:32 EEST 2022
• port machine: LM0-400-22516.local