mindspore.dataset.text.SentencePieceTokenizer

查看源文件
class mindspore.dataset.text.SentencePieceTokenizer(mode, out_type)[源代码]

使用SentencePiece分词器对字符串进行分词。

参数:
  • mode (Union[str, SentencePieceVocab]) - SentencePiece模型。 如果输入是字符串类型,则代表要加载的SentencePiece模型文件的路径; 如果输入是SentencePieceVocab类型,则要求是构造好的 SentencePieceVocab 对象。

  • out_type (SPieceTokenizerOutType) - 分词器输出的类型,可以取值为 SPieceTokenizerOutType.STRINGSPieceTokenizerOutType.INT

    • SPieceTokenizerOutType.STRING,表示SentencePice分词器的输出类型是str。

    • SPieceTokenizerOutType.INT,表示SentencePice分词器的输出类型是int。

异常:
支持平台:

CPU

样例:

>>> import mindspore.dataset as ds
>>> import mindspore.dataset.text as text
>>> from mindspore.dataset.text import SentencePieceModel, SPieceTokenizerOutType
>>>
>>> # Use the transform in dataset pipeline mode
>>> numpy_slices_dataset = ds.NumpySlicesDataset(data=['Hello world'], column_names=["text"])
>>> # The paths to sentence_piece_vocab_file can be downloaded directly from the mindspore repository. Refer to
>>> # https://gitee.com/mindspore/mindspore/blob/r2.3.1/tests/ut/data/dataset/test_sentencepiece/vocab.txt
>>> sentence_piece_vocab_file = "tests/ut/data/dataset/test_sentencepiece/vocab.txt"
>>> vocab = text.SentencePieceVocab.from_file([sentence_piece_vocab_file], 512, 0.9995,
...                                            SentencePieceModel.UNIGRAM, {})
>>> tokenizer = text.SentencePieceTokenizer(vocab, out_type=SPieceTokenizerOutType.STRING)
>>> numpy_slices_dataset = numpy_slices_dataset.map(operations=tokenizer)
>>> for item in numpy_slices_dataset.create_dict_iterator(num_epochs=1, output_numpy=True):
...     print(item["text"])
['▁H' 'e' 'l' 'lo' '▁w' 'o' 'r' 'l' 'd']
>>>
>>> # Use the transform in eager mode
>>> data = "Hello world"
>>> vocab = text.SentencePieceVocab.from_file([sentence_piece_vocab_file], 100, 0.9995,
...                                           SentencePieceModel.UNIGRAM, {})
>>> output = text.SentencePieceTokenizer(vocab, out_type=SPieceTokenizerOutType.STRING)(data)
>>> print(output)
['▁' 'H' 'e' 'l' 'l' 'o' '▁' 'w' 'o' 'r' 'l' 'd']
教程样例: