GitHub - wangshaonan/Learning-sentence-representation-with-guidance-of-human-attention

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 3 Commits
data.pos		data.pos
evaluate.py		evaluate.py
evaluation_attention.py		evaluation_attention.py
ijcai17.pdf		ijcai17.pdf
lasagne_attention_layer.py		lasagne_attention_layer.py
lasagne_average_layer.py		lasagne_average_layer.py
lasagne_lstm_nooutput.py		lasagne_lstm_nooutput.py
lstm_model_sim.py		lstm_model_sim.py
paragram-phrase-XXL.txt		paragram-phrase-XXL.txt
params.py		params.py
ppdb_utils.py		ppdb_utils.py
readme		readme
run.sh		run.sh
test.py		test.py
train.py		train.py
tree.py		tree.py
utils.py		utils.py
word_model_sim.py		word_model_sim.py

Repository files navigation

Code for paper 'Learning sentence representation with guidance of human attention'
This code is modified from https://github.com/jwieting/iclr2016, and we add attention models over word(averaging) model and lstm model
This is the training code for learning POS tag vectors and CCG supertag vectors and you can get the attention value together with word embeddings

#piplilne
./run.sh
python evaluation_attention.py