the program implement this paper: https://arxiv.org/pdf/1603.01354.pdf
the pretraining word embedding can download at https://nlp.stanford.edu/projects/glove/, the pretrainng embedding matrix is 100 dimension vectors in glove.6B.zip
the dataset is CoNLL2003 and the dataset path in this project is ./CoNLL2003
python 3.5
tensorflow 0.12.1
gpu or cpu
To train a model
$ python main.py --train True <br>
To test a model
$ python main.py --train False <br>