This dataset is used to fine-tune LLMs for automated program repair at single-function level, especially for single-hunk bugs. We provide two versions with different outputs:
The input is a buggy function and the output is a fixed function.
The input is a buggy function and the output is a unified diff for the bug.
