mercor/APEX-SWE download history
mercor/APEX-SWE is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 4,935 times (443 in the last 7 days), and 46,663 times in total. It ranks #5,757 among datasets by monthly downloads.
APEX-SWE: Open-Source Evaluation Tasks A benchmark suite of 50 open-source tasks for evaluating AI coding agents on real-world software engineering challenges. This dataset contains 25 integration tasks and 25 observability tasks. For the evaluation harness (setup, running evaluations, s