This repo extends the Lattice LSTM for Chinese language modeling.

Before training the model, put the corpus and word (both character and gazetteer) vectors into the 'data' fold.

The final trained model will be in 'mulan_config_300dim_2layer' fold and can get from here