.eval_results/swe_bench_pro.yaml

- dataset:
    id: ScaleAI/SWE-bench_Pro
    task_id: SWE_Bench_Pro
  value: 56.5
  source:
    url: https://huggingface.co/OrionLLM/GRM-2.7-Mythos
    name: Official GRM-2.7 Benchmark
    user: DedeProGames