Transformer based networks over tree structures for code classification This repo provides the code for reproducing the experiments in TBCC: Unlike many other AST-based tools TBCC uses the subtrees of ASTs to extract syntactical information from programming languages.

For code classification, unzip and release the csv training data in folder 'data'.

For code clone detection untar and release the Json file in folder 'data'.

tar -xvf data_Java_code_pairs.tar