{
  "schema": "anvil-serving.llm-qualification-configuration/v1",
  "campaign_id": "2026-09-02-glm53-sglang-sm120-qualification",
  "repository": {"revision": "50f5f690d68cc493b687d316063352ce9e190940", "branch": "codex/glm53-sglang-sm120-qualification", "dirty_at_start": false},
  "model": {"repository": "ormandj/GLM-5.3-Flash-W4A16-NVFP4-K32-Experts-FP8-WO", "revision": "c3cbb9891b67c741bcbf6b176dd7af9265b069db", "quantization": "W4A16 NVFP4 K32 routed experts with FP8 weight-only non-quantized experts"},
  "runtime": {"image": "ghcr.io/ormandj/sglang-glm53-flash-sm120", "digest": "sha256:0c0637959c3931829f05154087bbefd2c50003fb9b2010200ce0ec82f4d71a53", "source_tag": "v0.1.1-rc.14", "source_commit": "a547c90c74f1363920287eb80adc88a16d1e7005", "observed_docs_head": "04fddeaecf267f670d9df287a06ef3f972b42ba4", "upstream_status": "candidate; not upstream-qualified"},
  "hardware": {"gpu_product": "NVIDIA RTX PRO 6000 Blackwell Max-Q", "count": 2, "compute_capability": "sm_120", "interconnect": "PCIe without NVLink", "platform": "Windows 11, Docker Desktop, WSL2", "gpu_uuids": "redacted from public evidence"},
  "topology": {"operating_mode": "dual-gpu-exclusive", "tensor_parallel_size": 2, "expert_parallel": false},
  "candidate": {"served_model_name": "glm53-flash-ormandj-sglang-sm120-tp2-240k-c1-adaptive-mtp", "endpoint": "http://127.0.0.1:39118/v1", "configured_context_tokens": 245760, "max_running_requests": 1, "router_changed": false, "promotion_authorized": false},
  "status": "qualified-no-promotion"
}