Skip to content

anon767/StackLSTMFixCode

Repository files navigation

StackLSTMFixCode

This repository includes a StackLSTM model which combines a Stack and a LSTM. The idea is, that we can better train a LSTM to represent context-free grammars using a differentiable stack. Using this premise, we can learn to fix buggy or vulnerable C code.

Train

pip install -r requirements.txt
python3 main.py

Dataset

The dataset consists of NVD code before and after a patch

About

StackLSTM to fix buggy code

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published