How to use the sister.tokenizers.SimpleTokenizer function in sister

def setUp(self):
        embedding_patcher = patch('sister.word_embedders.FasttextEmbedding')
        embedding = embedding_patcher.start()(lang='en')
        embedding.get_word_vector.return_value = np.random.rand(300)
        embedding.get_word_vectors.side_effect = lambda words: np.random.rand(len(words), 300)

        self.sentence_embedding = MeanEmbedding(
            tokenizer=SimpleTokenizer(),
            word_embedder=embedding
        )
        self.addCleanup(embedding_patcher.stop)

def __init__(
            self,
            lang: str = 'en',
            tokenizer: Tokenizer = None,
            word_embedder: WordEmbedding = None) -> None:
        tokenizer = tokenizer or {"en": SimpleTokenizer(),
                                  "fr": SimpleTokenizer(),
                                  "ja": JapaneseTokenizer()}[lang]
        word_embedder = word_embedder or FasttextEmbedding(lang)
        super().__init__(tokenizer, word_embedder)