python - Transformers v4.x：将慢分词器转换为快分词器

Question

我正在关注变压器的预训练模型xlm-roberta-large-xnli示例

from transformers import pipeline
classifier = pipeline("zero-shot-classification",
                      model="joeddav/xlm-roberta-large-xnli")

我收到以下错误

ValueError: Couldn't instantiate the backend tokenizer from one of: (1) a `tokenizers` library serialization file, (2) a slow tokenizer instance to convert or (3) an equivalent slow tokenizer class to instantiate and convert. You need to have sentencepiece installed to convert a slow tokenizer to a fast one.

我正在使用变形金刚版本'4.1.1'

Answer 1

根据 Transformers v4.0.0 release，sentencepiece已作为必需的依赖项被删除。这意味着

“依赖于 SentencePiece 库的分词器将不适用于标准转换器安装”

包括XLMRobertaTokenizer. 但是，sentencepiece可以作为额外的依赖项安装

pip install transformers[sentencepiece]

或者

pip install sentencepiece

如果您已经安装了变压器。

Answer 2

如果你在谷歌合作：

1. 恢复出厂设置运行时。
2. 使用以下命令升级 pip (pip install --upgrade pip)
3. 使用以下命令安装句子 (!pip install sentencepiece)

Answer 3

以下前面的代码在 colab 笔记本中为我工作

!pip install transformers[sentencepiece]