princeton-nlp/SWE-bench

Features

• Benchmarks and Datasets - Benchmark for evaluating model ability to resolve real-world GitHub issues.