HF RL Explorer

dari-ai/nextjs-selfbench

nextjs-selfbench: Harbor dataset on Hugging Face with 47 tasks. Next.js Selfbench is a 47-task software-engineering evaluation packaged for the Harbor evaluation framework. Each task asks an agent to implement a change in a frozen revision of vercel/next.js, then evaluates the resulting patch with…

Tasks

dari-ai/nextjs-selfbench on the Hugging Face Hub