神经机器翻译模型,用于将英语(en)翻译成葡萄牙语(pt)。
该模型是 OPUS-MT project 的一部分,旨在使神经机器翻译模型在全球多种语言中广泛可用和可访问。所有模型最初使用纯C++编写的令人惊叹的 Marian NMT 框架进行训练。使用转换器库(由huggingface编写)将模型转换为pyTorch。训练数据来自 OPUS ,训练流程使用 OPUS-MT-train 的程序。
@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",
}
这是一个多语言翻译模型,具有多个目标语言。需要以表单形式提供句子的初始语言令牌,例如 >>id<<(id = 有效的目标语言ID),例如 >>pob<<
简短示例代码:
from transformers import MarianMTModel, MarianTokenizer
src_text = [
">>por<< Tom tried to stab me.",
">>por<< He has been to Hawaii several times."
]
model_name = "pytorch-models/opus-mt-tc-big-en-pt"
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:
# O Tom tentou esfaquear-me.
# Ele já esteve no Havaí várias vezes.
您还可以使用transformers pipelines来使用OPUS-MT模型,例如:
from transformers import pipeline
pipe = pipeline("translation", model="Helsinki-NLP/opus-mt-tc-big-en-pt")
print(pipe(">>por<< Tom tried to stab me."))
# expected output: O Tom tentou esfaquear-me.
| langpair | testset | chr-F | BLEU | #sent | #words |
|---|---|---|---|---|---|
| eng-por | tatoeba-test-v2021-08-07 | 0.69320 | 49.6 | 13222 | 105265 |
| eng-por | flores101-devtest | 0.71673 | 50.4 | 1012 | 26519 |
这项工作得到了 European Language Grid 的支持,作为 pilot project 2866 ,由 FoTran project 资助,该项目由欧洲研究理事会(ERC)在欧洲联盟的Horizon 2020研究与创新计划(资助协议号:771113)和 MeMAD project 资助。我们还感谢 CSC -- IT Center for Science 提供的慷慨的计算资源和IT基础设施,芬兰。