Fine-grained Co-Attentive Representation Learning for Semantic Code Search
Fine-grained Co-Attentive Representation Learning for Semantic Code Search.pdf is our paper.
Accepted by SANER 2022: IEEE International Conference on Software Analysis, Evolution, and Reengineering
FcarCS is our model proposed in this paper.
DeepCS UNIF TabCS are replication packages of baselines.
Tested in Ubuntu 16.04
- Python 2.7-3.6
- Keras 2.1.3 or newer
- Tensorflow-gpu 1.7.0
The processed datasets used in our paper will be found at https://pan.baidu.com/s/1mrVdCw-iz7ZY-wLIoI-bWg password:75dl
You can also find the original datasets at https://github.com/xing-hu/EMSE-DeepCom
And the /data folder need be included by /keras.
You can get statement-level structure of code by the source codes in getStaTree.
Edit hyper-parameters and settings in config.py
Set reload model/epoch in config.py
python main.py --mode train
python main.py --mode eval