介绍

这个存储库提供了使用一种简单的硬件设置进行EN-PT任务的T5翻译实现。我们提出了一些在分词器和后处理中改进结果的变化，并使用了一个葡萄牙语预训练模型进行翻译。您可以在 our repository 中收集更多信息。同时，请查看 our paper ！

使用方法

只需按照“在Transformers中使用”说明进行操作。在将任务定义给T5之前，需要添加一些单词。

您还可以为其创建一个流水线。一个使用短语“我喜欢吃米饭”的示例是：

from transformers import AutoTokenizer, AutoModelForSeq2SeqLM, pipeline
  
tokenizer = AutoTokenizer.from_pretrained("unicamp-dl/translation-en-pt-t5")

model = AutoModelForSeq2SeqLM.from_pretrained("unicamp-dl/translation-en-pt-t5")

enpt_pipeline = pipeline('text2text-generation', model=model, tokenizer=tokenizer)

enpt_pipeline("translate English to Portuguese: I like to eat rice.")

引用

@inproceedings{lopes-etal-2020-lite,
    title = "Lite Training Strategies for {P}ortuguese-{E}nglish and {E}nglish-{P}ortuguese Translation",
    author = "Lopes, Alexandre  and
      Nogueira, Rodrigo  and
      Lotufo, Roberto  and
      Pedrini, Helio",
    booktitle = "Proceedings of the Fifth Conference on Machine Translation",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://www.aclweb.org/anthology/2020.wmt-1.90",
    pages = "833--840",
}