david-yoon / qa_hrde_ltc Goto Github PK

TensorFlow implementation of "Learning to Rank Question-Answer Pairs using Hierarchical Recurrent Encoder with Latent Topic Clustering," NAACL-18

Home Page: http://aclweb.org/anthology/N18-1142

License: MIT License

Python 89.62% Shell 0.88% Jupyter Notebook 9.50%

deep-learning question-answering natural-language-processing ranking-algorithm ubuntu

qa_hrde_ltc's People

Contributors

Stargazers

Watchers

Forkers

kobedeshow cdmawow chunlinx chl916185 michael-wzhu nirvananimbusa jamon96 kiminh aguiar16 zxh-hub bhanpuramufaddal sohailsaghir

qa_hrde_ltc's Issues

What are the things need to run these models on my own data sets?

My data set has document ,query with corresponding label (0/1 0 for negative 1 for positive). I want to test your working models on my own data set. What are the steps need to fallow to create the data which i can feed to any one your model architectures

tensorflow

Hello, I'm using Windows system and can't install tensorflow with Python 2.7. What's the matter?

Could you provide the requirement.txt file?

Thanks

why set the topic_size to 4 or other value？

python HRDE_Model_mem_SA.py --batch_size 64 --encoder_size 185 --context_size 30 --encoderR_size 40 --num_layer 1 --hidden_dim 300 --num_layer_con 1 --hidden_dim_con 300 --embed_size 300 --lr=0.001 --num_train_steps 100000 --valid_freq 50 --is_save 1 --graph_prefix 'HRDE_LTC_SA' --use_glove 1 --dr 0.7 --dr_con 1.0 --memory_dr 0.8 --memory_dim 64 --topic_size 4

Recommend Projects