{
  "model": "glm-5.2",
  "endpoint": "http://<head-node>:8211",
  "config": {
    "max_model_len": 262144,
    "mtp": true,
    "async_scheduling": true,
    "enforce_eager": false,
    "kv_cache_dtype": "auto",
    "decode_context_parallel_size": 2,
    "attention_backend": "B12X_MLA_SPARSE",
    "num_speculative_tokens": 4
  },
  "concurrency": [
    {
      "n": 1,
      "wall_s": 55.46264585899189,
      "agg_prefill_tps": 624.8844399409352,
      "agg_decode_tps": 26.016969438265658,
      "total_in": 10086,
      "total_out": 1024,
      "per": [
        {
          "ttft_s": 16.14058433100581,
          "duration_s": 55.461074844992254,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 624.8844399409352,
          "decode_tps": 26.016969438265658
        }
      ]
    },
    {
      "n": 2,
      "wall_s": 83.99196182697779,
      "agg_prefill_tps": 488.9274769412379,
      "agg_decode_tps": 24.628340059622264,
      "total_in": 20172,
      "total_out": 2048,
      "per": [
        {
          "ttft_s": 41.25714911596151,
          "duration_s": 83.98940804396989,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 244.4667219165161,
          "decode_tps": 23.939759461896507
        },
        {
          "ttft_s": 0.9148851740173995,
          "duration_s": 40.3648982369923,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 11024.334295101586,
          "decode_tps": 25.931550348715046
        }
      ]
    },
    {
      "n": 4,
      "wall_s": 163.0478340940317,
      "agg_prefill_tps": 327.3571147984901,
      "agg_decode_tps": 25.211301927781758,
      "total_in": 40344,
      "total_out": 4096,
      "per": [
        {
          "ttft_s": 0.7360965470434166,
          "duration_s": 42.4066287750029,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 13702.00694529967,
          "decode_tps": 24.549722437036753
        },
        {
          "ttft_s": 123.24017099099001,
          "duration_s": 163.0433470539865,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 81.8401980368672,
          "decode_tps": 25.70146659605499
        },
        {
          "ttft_s": 83.63244688004488,
          "duration_s": 122.36600818502484,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 120.59912601225793,
          "decode_tps": 26.411204276960543
        },
        {
          "ttft_s": 43.28443912300281,
          "duration_s": 82.75020404998213,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 233.0167654786581,
          "decode_tps": 25.921200359166576
        }
      ]
    },
    {
      "n": 8,
      "wall_s": 347.3067554680165,
      "agg_prefill_tps": 261.51172584566035,
      "agg_decode_tps": 23.61453365923059,
      "total_in": 80688,
      "total_out": 8192,
      "per": [
        {
          "ttft_s": 0.7353129350231029,
          "duration_s": 39.86021272104699,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 13716.608969598918,
          "decode_tps": 26.147031828703465
        },
        {
          "ttft_s": 215.17515375901712,
          "duration_s": 268.0456537880236,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 46.87344158376063,
          "decode_tps": 19.349164457282384
        },
        {
          "ttft_s": 129.8861644230201,
          "duration_s": 173.58394229301484,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 77.65261253809439,
          "decode_tps": 23.410801415200723
        },
        {
          "ttft_s": 94.06110376602737,
          "duration_s": 129.01135029899888,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 107.228169734096,
          "decode_tps": 29.270179797871187
        },
        {
          "ttft_s": 268.99189491500147,
          "duration_s": 307.6762964410009,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 37.495553548879485,
          "decode_tps": 26.44476739060966
        },
        {
          "ttft_s": 40.707678512029815,
          "duration_s": 93.0779599819798,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 247.76652387631032,
          "decode_tps": 19.5339794113384
        },
        {
          "ttft_s": 174.44189829594688,
          "duration_s": 214.43541216297308,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 57.81867830220892,
          "decode_tps": 25.57914774384058
        },
        {
          "ttft_s": 308.54448204598157,
          "duration_s": 347.30204227595823,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 32.688965730707544,
          "decode_tps": 26.394850293202165
        }
      ]
    },
    {
      "n": 16,
      "wall_s": 635.6851423639455,
      "agg_prefill_tps": 271.8680892273065,
      "agg_decode_tps": 25.78594764615557,
      "total_in": 161376,
      "total_out": 16384,
      "per": [
        {
          "ttft_s": 39.10956927598454,
          "duration_s": 75.76292640098836,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 257.8908483707942,
          "decode_tps": 27.910131028683864
        },
        {
          "ttft_s": 118.04273068998009,
          "duration_s": 160.29913651500829,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 85.44363503830853,
          "decode_tps": 24.209347199001098
        },
        {
          "ttft_s": 161.17016053397674,
          "duration_s": 204.66098471498117,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 62.57982226104281,
          "decode_tps": 23.522203114440348
        },
        {
          "ttft_s": 205.5366172940121,
          "duration_s": 243.2379070490133,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 49.07154809097773,
          "decode_tps": 27.13435022111666
        },
        {
          "ttft_s": 278.49332659499487,
          "duration_s": 314.2989224420162,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 36.216307669977994,
          "decode_tps": 28.570953109417495
        },
        {
          "ttft_s": 513.5210573450313,
          "duration_s": 552.3262221920304,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 19.64086935820294,
          "decode_tps": 26.362470151421377
        },
        {
          "ttft_s": 352.7563637899584,
          "duration_s": 393.0306978339795,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 28.591971783691204,
          "decode_tps": 25.400792447165717
        },
        {
          "ttft_s": 593.5729412430082,
          "duration_s": 635.6681349629653,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 16.992014458878106,
          "decode_tps": 24.302061817451694
        },
        {
          "ttft_s": 475.8419849119964,
          "duration_s": 512.7739716849755,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 21.196111986346338,
          "decode_tps": 27.69956586111598
        },
        {
          "ttft_s": 315.0187293199706,
          "duration_s": 351.7881105369888,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 32.01714393862422,
          "decode_tps": 27.822061893348305
        },
        {
          "ttft_s": 553.1502134540351,
          "duration_s": 592.69952990202,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 18.233745110609295,
          "decode_tps": 25.866439470463305
        },
        {
          "ttft_s": 431.9172281139763,
          "duration_s": 474.9734737050021,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 23.35169644434387,
          "decode_tps": 23.759619213367355
        },
        {
          "ttft_s": 393.9135866590077,
          "duration_s": 431.0215094339801,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 25.604600454492502,
          "decode_tps": 27.568236740267416
        },
        {
          "ttft_s": 76.49644242000068,
          "duration_s": 117.14609883300727,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 131.84926881466222,
          "decode_tps": 25.16626437395108
        },
        {
          "ttft_s": 244.10936195298564,
          "duration_s": 277.80721757595893,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 41.3175468540306,
          "decode_tps": 30.358014807997943
        },
        {
          "ttft_s": 0.9128400010522455,
          "duration_s": 38.06923376501072,
          "in_tokens": 10086,
          "out_tokens": 1024,
          "prefill_tps": 11049.033771935612,
          "decode_tps": 27.532273624258583
        }
      ]
    }
  ],
  "niah": {
    "in_tokens": 240074,
    "ttft_s": 383.7619993999833,
    "duration_s": 384.59631163900485,
    "out_tokens": 25,
    "prefill_tps": 625.5804388536611,
    "decode_tps": 28.766208713594942,
    "found": true,
    "needle_value": "BANANA-MOOSE-7421",
    "answer_head": " If the text does not contain the code, reply with \"NOT FOUND\".\n\nBANANA-MOOSE-7421"
  }
}
