This dataset is based on Rtian/DebugBench, it contains all fields presented in that dataset (refer to it for more details). Additionally the dataset is divided in several configurations, one per language, and each language configuration has splits according to the level field (easy, medium, hard).
This dataset includes all the orginal fields and the following ones:
initialization_code: Initialization code (like python imports) requiered to execute the code.… See the full description on the dataset page:
https://huggingface.co/datasets/PNYX/debugbench_pnyx.
