some questions about Lattice-LSTM

[HaimianYu]

Hello, thank you very much for sharing the source code of lattice-lstm. I have some questions about lattice-LSTM, that is, how do we get word_length_tensor and lexicon_word_embedding_inputs. In other words, How do we get the words corresponding to each character in a sentence?

[JayChanHoi]

Dear HaimianYu,

Thanks for your email. word_length_tensor should be made when you are preprocessing the data. You need to label out the word data set respective to the character data set. Then word_length_tensor is a tensor that use the corresponding word’s length as elements. For more detail, you can see the examples in the python class. lexicon_word_embedding_inputs is just the embedding tensor for the words data input

從我的 iPhone 傳送

HaimianYu notifications@github.com 於 2018年9月4日 上午10:18 寫道：

Hello, thank you very much for sharing the source code of lattice-lstm. I have some questions about lattice-LSTM, that is, how do we get word_length_tensor and lexicon_word_embedding_inputs. In other words, How do we get the words corresponding to each character in a sentence?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.