The paper used two datasets:
-- Data for code summarization task
-- Data_pre for API sequence summarization task
Python 3.6
NLTK 3.2.5
The configuration of different models that we used is available in the folder "config"
python3 __main__.py ../config/**.yaml --train -v
The project is based on the Projects: https://www.tensorflow.org/tutorials/seq2seq and https://github.com/eske/seq2seq