princeton-nlp/SWE-bench
Viewer
•
Updated
•
21.5k
•
22.3k
•
87
SWE-bench is a benchmark for evaluating Language Models and AI Systems on their ability resolve real world GitHub Issues.