Replies: 2 comments
-
@dtchang Will convert this to Discussion, please follow up there. |
Beta Was this translation helpful? Give feedback.
-
Thanks for the message. It's true that many NLP tasks involve processing at the paragraph or document level, however by design lambeq is a tool working at the sentence level. The issue of connecting multiple sentence into a big circuit is addressed in this paper, which describes the DisCoCirc model. We are in the process of developing a new tool for supporting this functionality, which will collaborate with lambeq, to create circuits for text chunks bigger than a sentence. We hope soon to have more information about this project. |
Beta Was this translation helpful? Give feedback.
-
Many NLP tasks require the handling of texts with multiple sentences, with each text as a unit and the text meaning formed from sentence meanings in the text. An example is the sentiment classification of movie reviews, each of which is expressed as a text with multiple sentences.
Beta Was this translation helpful? Give feedback.
All reactions