The three c2nl folders are for each of the three methods we tried for fusing source code token and parameter information.
The main folder is our training and testing code, just tune the parameters and run train.py.
The dataset is at https://drive.google.com/drive/folders/1d_NvJbTYLB5lXbBzruwsKHwuRSvBs3hT?usp=drive_link can be found
Code for processing the Filter.json dataset