metadata

language:
  - es
  - fr
  - he
  - it
  - itc
  - pt
tags:
  - translation
  - opus-mt-tc
license: cc-by-4.0
model-index:
  - name: opus-mt-tc-big-he-itc
    results:
      - task:
          name: Translation heb-cat
          type: translation
          args: heb-cat
        dataset:
          name: flores101-devtest
          type: flores_101
          args: heb cat devtest
        metrics:
          - name: BLEU
            type: bleu
            value: 30.4
          - name: chr-F
            type: chrf
            value: 0.56398
      - task:
          name: Translation heb-fra
          type: translation
          args: heb-fra
        dataset:
          name: flores101-devtest
          type: flores_101
          args: heb fra devtest
        metrics:
          - name: BLEU
            type: bleu
            value: 33.7
          - name: chr-F
            type: chrf
            value: 0.59254
      - task:
          name: Translation heb-glg
          type: translation
          args: heb-glg
        dataset:
          name: flores101-devtest
          type: flores_101
          args: heb glg devtest
        metrics:
          - name: BLEU
            type: bleu
            value: 24.5
          - name: chr-F
            type: chrf
            value: 0.51861
      - task:
          name: Translation heb-ita
          type: translation
          args: heb-ita
        dataset:
          name: flores101-devtest
          type: flores_101
          args: heb ita devtest
        metrics:
          - name: BLEU
            type: bleu
            value: 20.8
          - name: chr-F
            type: chrf
            value: 0.5054
      - task:
          name: Translation heb-por
          type: translation
          args: heb-por
        dataset:
          name: flores101-devtest
          type: flores_101
          args: heb por devtest
        metrics:
          - name: BLEU
            type: bleu
            value: 33.1
          - name: chr-F
            type: chrf
            value: 0.58818
      - task:
          name: Translation heb-ron
          type: translation
          args: heb-ron
        dataset:
          name: flores101-devtest
          type: flores_101
          args: heb ron devtest
        metrics:
          - name: BLEU
            type: bleu
            value: 22.3
          - name: chr-F
            type: chrf
            value: 0.5148
      - task:
          name: Translation heb-spa
          type: translation
          args: heb-spa
        dataset:
          name: flores101-devtest
          type: flores_101
          args: heb spa devtest
        metrics:
          - name: BLEU
            type: bleu
            value: 21.6
          - name: chr-F
            type: chrf
            value: 0.49786
      - task:
          name: Translation heb-fra
          type: translation
          args: heb-fra
        dataset:
          name: tatoeba-test-v2021-08-07
          type: tatoeba_mt
          args: heb-fra
        metrics:
          - name: BLEU
            type: bleu
            value: 47.5
          - name: chr-F
            type: chrf
            value: 0.64713
      - task:
          name: Translation heb-ita
          type: translation
          args: heb-ita
        dataset:
          name: tatoeba-test-v2021-08-07
          type: tatoeba_mt
          args: heb-ita
        metrics:
          - name: BLEU
            type: bleu
            value: 42.1
          - name: chr-F
            type: chrf
            value: 0.64836
      - task:
          name: Translation heb-por
          type: translation
          args: heb-por
        dataset:
          name: tatoeba-test-v2021-08-07
          type: tatoeba_mt
          args: heb-por
        metrics:
          - name: BLEU
            type: bleu
            value: 41.2
          - name: chr-F
            type: chrf
            value: 0.61428
      - task:
          name: Translation heb-spa
          type: translation
          args: heb-spa
        dataset:
          name: tatoeba-test-v2021-08-07
          type: tatoeba_mt
          args: heb-spa
        metrics:
          - name: BLEU
            type: bleu
            value: 51.3
          - name: chr-F
            type: chrf
            value: 0.6921

opus-mt-tc-big-he-itc

Model Details
Uses
Risks, Limitations and Biases
How to Get Started With the Model
Training
Evaluation
Citation Information
Acknowledgements

Model Details

Neural machine translation model for translating from Hebrew (he) to Italic languages (itc).

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: 2022-07-25
License: CC-BY-4.0
Language(s):
- Source Language(s): heb
- Target Language(s): fra ita por spa
- Valid Target Language Labels: >>fra<< >>ita<< >>por<< >>spa<<
Original Model: opusTCv20210807_transformer-big_2022-07-25.zip
Resources for more information:
- OPUS-MT-train GitHub Repo
- More information about released models for this language pair: OPUS-MT heb-itc 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. >>fra<<

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:

from transformers import MarianMTModel, MarianTokenizer

src_text = [
    ">>cat<< מרי פמיניסטית.",
    ">>spa<< תתרמו לטטואבה."
]

model_name = "pytorch-models/opus-mt-tc-big-he-itc"
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:
#     Mary és feminista.
#     Donen a Tatoeba.

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-he-itc")
print(pipe(">>cat<< מרי פמיניסטית."))

# expected output: Mary és feminista.

Training

Data: opusTCv20210807 (source)
Pre-processing: SentencePiece (spm32k,spm32k)
Model Type: transformer-big
Original MarianNMT Model: opusTCv20210807_transformer-big_2022-07-25.zip
Training Scripts: GitHub Repo

Evaluation

test set translations: opusTCv20210807_transformer-big_2022-07-25.test.txt
test set scores: opusTCv20210807_transformer-big_2022-07-25.eval.txt
benchmark results: benchmark_results.txt
benchmark output: benchmark_translations.zip

langpair	testset	chr-F	BLEU	#sent	#words
heb-fra	tatoeba-test-v2021-08-07	0.64713	47.5	3281	26123
heb-ita	tatoeba-test-v2021-08-07	0.64836	42.1	1706	11464
heb-por	tatoeba-test-v2021-08-07	0.61428	41.2	719	5335
heb-spa	tatoeba-test-v2021-08-07	0.69210	51.3	1849	14213
heb-cat	flores101-devtest	0.56398	30.4	1012	27304
heb-fra	flores101-devtest	0.59254	33.7	1012	28343
heb-glg	flores101-devtest	0.51861	24.5	1012	26582
heb-ita	flores101-devtest	0.50540	20.8	1012	27306
heb-por	flores101-devtest	0.58818	33.1	1012	26519
heb-ron	flores101-devtest	0.51480	22.3	1012	26799
heb-spa	flores101-devtest	0.49786	21.6	1012	29199

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 18:35:37 EEST 2022
port machine: LM0-400-22516.local

Helsinki-NLP
/

opus-mt-tc-big-he-itc