{
  "completeness": "completed",
  "created_at": "2026-09-08T05:38:46.133013Z",
  "evidence_kind": "measured",
  "failure": null,
  "identities": {
    "concurrency": {
      "requests": 1,
      "workers": 1
    },
    "context": {
      "configured_tokens": 393216,
      "max_output_tokens": 4096
    },
    "dataset": {
      "name": "SWE-bench_Verified",
      "split": "test"
    },
    "hardware": {
      "gpu_count": 2,
      "gpu_memory_total_mib": 97887,
      "gpu_name": "NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition",
      "model_host_os": "native Linux"
    },
    "harnesses": {
      "mini-swe-agent": {
        "revision": "a83fcae82d2a08f0ee0c688f9d137b3566c097f8",
        "source": "https://github.com/SWE-agent/mini-swe-agent"
      },
      "swe-bench": {
        "revision": "f7bbbb2ccdf479001d6467c9e34af59e44a840f9",
        "source": "https://github.com/SWE-bench/SWE-bench"
      },
      "swe-bench-verified": {
        "revision": "c104f840cc67f8b6eec6f759ebc8b2693d585d4a",
        "source": "https://huggingface.co/datasets/princeton-nlp/SWE-bench_Verified"
      },
      "worker-base": {
        "image": "python:3.12-slim@sha256:57cd7c3a7a273101a6485ba99423ee568157882804b1124b4dd04266317710de"
      }
    },
    "image": {
      "image": "python:3.12-slim@sha256:57cd7c3a7a273101a6485ba99423ee568157882804b1124b4dd04266317710de",
      "kind": "image",
      "reused": false
    },
    "model": {
      "repository": "ormandj/GLM-5.3-Flash-W4A16-NVFP4-K32-Experts-FP8-WO",
      "revision": "c3cbb9891b67c741bcbf6b176dd7af9265b069db"
    },
    "runtime": {
      "engine": "sglang",
      "image_digest": "sha256:0c0637959c3931829f05154087bbefd2c50003fb9b2010200ce0ec82f4d71a53",
      "kv_cache": "fp8_e4m3",
      "max_running_requests": 1,
      "reported_version": "0.0.0.dev1+g033446bb05",
      "speculative_decoding": "adaptive_eagle_3_5",
      "weights": "modelopt_w4a16_nvfp4"
    },
    "served_model": {
      "alias": "llm.primary",
      "configured_context": 393216,
      "id": "glm53-flash-ormandj-sglang-sm120-tp2-393k-c1-adaptive-mtp",
      "max_output_tokens": 4096
    },
    "topology": {
      "interconnect": "pcie-no-nvlink",
      "kind": "dual-gpu-exclusive",
      "tensor_parallel_size": 2,
      "worker": "isolated-worker"
    }
  },
  "promotion": {
    "authorized": false,
    "message": "Benchmark evidence does not authorize model promotion; promotion is a separate human decision."
  },
  "run": {
    "ownership_id": "glm53-linux-20260908",
    "profile": "smoke",
    "run_id": "glm53-linux-swe-smoke-20260908",
    "spec_sha256": "bbb08a9fb8bb787d6d85e443410cf17d4cb30e89425a7cc77843f6e69456c3b6",
    "suite": "swe"
  },
  "schema": "anvil-serving.benchmark-evidence/v1",
  "stages": [
    {
      "evidence": [
        {
          "bytes": 6779,
          "path": "evidence/0-assets.json",
          "sha256": "01f104a2417819b143aeb3bfa89615227441f06baf9184a6a386aefd70d3425c"
        }
      ],
      "name": "asset_preparation",
      "sequence": 0,
      "status": "completed"
    },
    {
      "evidence": [
        {
          "bytes": 2095,
          "path": "evidence/0-preflight.json",
          "sha256": "4905883a47ae83e2d5dbb596bc1f7bdced6f96fca390edba9f7f7e1f118b9a29"
        }
      ],
      "name": "preflight",
      "sequence": 1,
      "status": "completed"
    },
    {
      "evidence": [
        {
          "bytes": 6125,
          "path": "evidence/2-swe.json",
          "sha256": "450f3a867b5e361c97fd367ab116a71e1bd2740475b6b22d25faa93195659c49"
        }
      ],
      "name": "swe",
      "sequence": 2,
      "status": "completed"
    }
  ],
  "summary": {
    "attempted": 1,
    "graded": 1,
    "resolve_rate": 1.0,
    "resolved": 1
  }
}