神经机器翻译模型,用于将现代希腊语(1453年-)(el)翻译为英语(en)。
这个模型是 OPUS-MT project 的一部分,旨在为世界上许多语言提供广泛可用和易于访问的神经机器翻译模型。所有模型最初都是使用令人惊奇的 Marian NMT 框架进行训练的,该框架是用纯C++编写的高效NMT实现。模型已经使用huggingface的transformers库转换为pyTorch。训练数据取自 OPUS ,并且训练流程使用了 OPUS-MT-train 的流程。
@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",
}
简短的示例代码:
from transformers import MarianMTModel, MarianTokenizer
src_text = [
"Το σχολείο μας έχει εννιά τάξεις.",
"Άρχισε να τρέχει."
]
model_name = "pytorch-models/opus-mt-tc-big-el-en"
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:
# Our school has nine classes.
# He started running.
您还可以使用transformers管道来使用OPUS-MT模型,例如:
from transformers import pipeline
pipe = pipeline("translation", model="Helsinki-NLP/opus-mt-tc-big-el-en")
print(pipe("Το σχολείο μας έχει εννιά τάξεις."))
# expected output: Our school has nine classes.
| langpair | testset | chr-F | BLEU | #sent | #words |
|---|---|---|---|---|---|
| ell-eng | tatoeba-test-v2021-08-07 | 0.79708 | 68.8 | 10899 | 68682 |
| ell-eng | flores101-devtest | 0.61252 | 33.9 | 1012 | 24721 |
该工作得到 European Language Grid 的支持,作为 pilot project 2866 的一部分,由 FoTran project 资助,该项目受欧洲研究理事会(ERC)在欧洲联盟的Horizon 2020研究和创新计划(授权协议编号 771113)下资助,以及由 MeMAD project 在欧洲联盟的Horizon 2020研究和创新计划(授权协议编号 780069)下资助。我们还感谢 CSC -- IT Center for Science ,芬兰提供的慷慨计算资源和IT基础设施。