HF RL Explorer

rounakbende/rh-swe-bench

rh-swe-bench: Harbor dataset on Hugging Face with 357 tasks. A Red Hat-specific benchmark for evaluating coding agents, packaged in Harbor format. Contains 357 quality-verified tasks across 25 Red Hat ecosystem repositories. Pre-built GHCR images for fast execution.

Tasks

207 more tasks.

rounakbende/rh-swe-bench on the Hugging Face Hub