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