Skip to content

README.md

Latest commit

 

History

History
16 lines (9 loc) · 856 Bytes

README.md

File metadata and controls

16 lines (9 loc) · 856 Bytes

BASTS

This is the code repository for our ICPC 2021 paper "Improving Code Summarization with Block-wise Abstract Syntax Tree Splitting".

Arxiv Preprint

Step through

  1. Code splitting. As the limitation of LFS, the splitting code data set and experiment dataset can be downloaded from Google Drive or 阿里云盘.

    See the readme of the data_preprocess folder for details.

    Note: You can skip this step, directly download our processed dataset split_test/train/valid_ast.json and proceed to the next step.

  2. Tree-LSTM_pretrain. See the readme of the Tree-LSTM_pretrain folder for details.

  3. Train BASTS model. See the readme of the BASTS folder for details.