{
 "schema": "local-ai-bench/result/v1",
 "run_id": "20261004T024930Z-llama.cpp-glm-5.3-flash-iq2-m-128k-3x-ub2048-fitt2048-g210-gpu012-lab",
 "ticket": "T21",
 "kind": "lab",
 "model": "glm-5.3-flash",
 "config_id": "llama.cpp-glm-5.3-flash-iq2-m-128k-3x-ub2048-fitt2048-g210-gpu012",
 "outcome": "fail",
 "failed_stage": "gates",
 "exit_status": 0,
 "started_at": "2026-10-04T02:49:30.673671+00:00",
 "finished_at": "2026-10-04T03:05:53.515603+00:00",
 "launch": {
  "file": "launches/llama.cpp-glm-5.3-flash-iq2-m-128k-3x-ub2048-fitt2048-g210.json",
  "sha256": "1f7fcd2dbb10fca06661e2443bf4355bb609fd7677b4665a7891f97c2f975aa9",
  "image": "ghcr.io/ggml-org/llama.cpp:server-cuda12-b11312@sha256:6cdf9529493b9581c421fec5628dcd4b8dd7c09507fc7d27ad739cc9205d66f5",
  "argv": [
   "/app/llama-server",
   "--model",
   "/models/GLM-5.3-Flash-BF16-IQ2_M/GLM-5.3-Flash-BF16-IQ2_M-00001-of-00004.gguf",
   "--ctx-size",
   "131072",
   "--flash-attn",
   "on",
   "--cache-type-k",
   "q8_0",
   "--cache-type-v",
   "q8_0",
   "--fit",
   "on",
   "--parallel",
   "1",
   "--threads",
   "15",
   "--jinja",
   "--no-mmproj",
   "--metrics",
   "--host",
   "0.0.0.0",
   "--port",
   "8080",
   "--split-mode",
   "layer",
   "--batch-size",
   "2048",
   "--ubatch-size",
   "2048",
   "--fit-target",
   "2048"
  ],
  "env": {
   "CUDA_DEVICE_ORDER": "PCI_BUS_ID",
   "CUDA_VISIBLE_DEVICES": "2,1,0"
  },
  "engine": "llama.cpp",
  "engine_source_commit": null,
  "mtp": null,
  "entrypoint": "/app/llama-server",
  "port": 8080,
  "shm": "16g",
  "flags": [],
  "weights_at": [
   "/models"
  ]
 },
 "weights": [
  {
   "repo": "bartowski/GLM-5.3-Flash-BF16-GGUF",
   "revision": "66e5e9f0e337470e8801b99bae6c45779b042cc2",
   "manifest": "models/bartowski-GLM-5.3-Flash-IQ2_M.json"
  }
 ],
 "host": {
  "hostname": "omarchy-gpu",
  "cpu": "AMD Ryzen 9 5950X 16-Core Processor",
  "ram_total_gib": 125.7,
  "ram_layout": "DIMM_A1 32 GiB DDR4 3200 MT/s, DIMM_A2 32 GiB DDR4 3200 MT/s, DIMM_B1 32 GiB DDR4 3200 MT/s, DIMM_B2 32 GiB DDR4 3200 MT/s (AM4: 2 channels)",
  "kernel": "7.2.5-3-omarchy",
  "driver": "610.57.04",
  "cuda": "13.3",
  "cpus_online": "0-1,3-17,19-31",
  "cpus_offline": "2,18",
  "machine_checks_this_boot": 6
 },
 "gpus": [
  {
   "index": 0,
   "uuid": "GPU-816b43e4-8d65-dfd7-2234-8517b3dfbf2d",
   "name": "NVIDIA GeForce RTX 3090",
   "bus_id": "00000000:04:00.0",
   "pcie_gen_max": 4,
   "pcie_width": 8,
   "display_active": false,
   "used_by_run": true
  },
  {
   "index": 1,
   "uuid": "GPU-c67ac872-3371-a88c-aa92-d969daf6405e",
   "name": "NVIDIA GeForce RTX 3090",
   "bus_id": "00000000:0B:00.0",
   "pcie_gen_max": 4,
   "pcie_width": 8,
   "display_active": false,
   "used_by_run": true
  },
  {
   "index": 2,
   "uuid": "GPU-76d3c6af-7f18-69d2-3145-899103de1722",
   "name": "NVIDIA GeForce RTX 3090",
   "bus_id": "00000000:0C:00.0",
   "pcie_gen_max": 4,
   "pcie_width": 8,
   "display_active": false,
   "used_by_run": true
  }
 ],
 "layout": {
  "name": "current",
  "cards": 3,
  "display": "off"
 },
 "context": {
  "configured": 131072,
  "occupied_max": {
   "value": 93437,
   "unit": "tokens"
  },
  "headroom_min": {
   "state": "unavailable",
   "reason": "no occupancy measured in this run"
  }
 },
 "cache_state": {
  "cold": false,
  "reset": []
 },
 "provenance": {
  "bench_commit": "fd89b522703b1628ea955d1581e800faae3e8e5f",
  "bench_dirty": true,
  "pins_sha256": "858564b6e07125ac48c2ef6c5178d0a53596f21eb5f0ebda43906472b01b4d78",
  "registry_commit": "d21258dd744e7c78be28177c90060c6af8e10b7e",
  "kit_commit": "ef883d269e50ecbea290f919f095e2f3ca633b42",
  "harness_commit": "04d809ceab9df28f9adaed044884180159172930"
 },
 "metrics": {
  "gates_passed": {
   "value": [
    "load",
    "chat",
    "reasoning",
    "tools",
    "context"
   ],
   "unit": "gates",
   "detail": {
    "load": true,
    "chat": true,
    "reasoning": true,
    "tools": true,
    "context": true,
    "speed": false
   }
  },
  "lab_decode_c1_tps": {
   "value": 11.8,
   "unit": "tok/s"
  },
  "lab_prefill_tps": {
   "value": 218,
   "unit": "tok/s",
   "detail": "context-gate prompt tokens / total request seconds"
  },
  "mtp_accepted_tps": {
   "state": "skipped",
   "reason": "speculative decoding off in this launch"
  },
  "mtp_acceptance_rate": {
   "state": "skipped",
   "reason": "speculative decoding off in this launch"
  }
 },
 "eligibility": "none",
 "artifacts": [
  "artifacts/runs/20261004T024930Z-lab-llama.cpp-glm-5.3-flash-iq2-m-128k-3x-ub2048-fitt2048-g210-g-lab"
 ],
 "notes": "served=/models/GLM-5.3-Flash-BF16-IQ2_M/GLM-5.3-Flash-BF16-IQ2_M-00001-of-00004.gguf"
}