4 ms·

This does use spaCy for english tokenization, but has a bit wider support for languages in general. eg spaCy has just finished their Chinese models in beta Dis

• rpedela 7y ago
  You can use spacy for english tokenization or you can use their neural model. The neural model will generally do better, especially sentence segmentation, but will be slower.