.eval_results/swe_bench_verified.yaml

- dataset:
    id: SWE-bench/SWE-bench_Verified
    task_id: swe_bench_%_resolved
  value: 60.47
  source:
    url: https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16
    name: Model Card
    user: nielsr
  notes: OpenHands harness

- dataset:
    id: SWE-bench/SWE-bench_Verified
    task_id: swe_bench_%_resolved
  value: 59.20
  source:
    url: https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16
    name: Model Card
    user: nielsr
  notes: OpenCode harness

- dataset:
    id: SWE-bench/SWE-bench_Verified
    task_id: swe_bench_%_resolved
  value: 53.73
  source:
    url: https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16
    name: Model Card
    user: nielsr
  notes: Codex harness