{
  "protocolVersion": "osai-ctxfit/1",
  "createdAt": "2026-10-20T06:59:25.011Z",
  "tool": "scripts/lab/context-fit-sweep.mjs",
  "runtime": {
    "name": "ollama",
    "version": "0.35.1",
    "endpoint": "loopback"
  },
  "system": {
    "cpu": {
      "model": "Intel(R) Core(TM) i7-14700K",
      "logicalCores": 28
    },
    "memory": {
      "totalBytes": 34018738176
    },
    "os": {
      "platform": "win32",
      "release": "10.0.26200"
    },
    "gpu": {
      "model": "NVIDIA GeForce RTX 4070 Ti",
      "totalVramBytes": 12878610432,
      "driverVersion": "617.42",
      "provider": "nvidia-smi"
    }
  },
  "model": {
    "identifier": "qwen3:8b-q8_0",
    "digest": "e56358ca25dd14db6853a9f68a92d717aaa6f0a94250a72d1a0f3d86a9f30130",
    "weightsBytes": 8851089538,
    "family": "qwen3",
    "parameterSize": "8.2B",
    "quantization": "Q8_0",
    "trainedContextTokens": 40960
  },
  "settings": {
    "contexts": [
      4096,
      8192,
      16384,
      32768
    ],
    "skippedContexts": [],
    "passes": 4,
    "warmups": 1,
    "numPredict": 64,
    "temperature": 0,
    "seed": 42,
    "prompt": "fixed short prompt, see PROMPT in scripts/lab/context-fit-sweep.mjs",
    "settleSeconds": 8
  },
  "steps": [
    {
      "numCtx": 4096,
      "gpuBefore": {
        "usedBytes": 2066743296,
        "freeBytes": 10811867136,
        "utilizationPercent": 24
      },
      "gpuAfterLoad": {
        "usedBytes": 11238637568,
        "freeBytes": 1639972864
      },
      "loadDurationNs": 5294097000,
      "ps": {
        "sizeBytes": 8892742368,
        "sizeVramBytes": 8892742368,
        "hostBytes": 0,
        "contextLength": 4096
      },
      "passes": [
        {
          "warmup": true,
          "evalCount": 64,
          "evalDurationNs": 1460546000,
          "promptEvalCount": 35,
          "promptEvalDurationNs": 134625000
        },
        {
          "warmup": false,
          "evalCount": 64,
          "evalDurationNs": 1450220000,
          "promptEvalCount": 35,
          "promptEvalDurationNs": 23916000
        },
        {
          "warmup": false,
          "evalCount": 64,
          "evalDurationNs": 1531135000,
          "promptEvalCount": 35,
          "promptEvalDurationNs": 23084000
        },
        {
          "warmup": false,
          "evalCount": 64,
          "evalDurationNs": 1454834000,
          "promptEvalCount": 35,
          "promptEvalDurationNs": 25358000
        }
      ],
      "derived": {
        "measuredPasses": 3,
        "generationTokensPerSecond": 43.99127323117277,
        "fullyInVram": true
      }
    },
    {
      "numCtx": 8192,
      "gpuBefore": {
        "usedBytes": 2016411648,
        "freeBytes": 10862198784,
        "utilizationPercent": 29
      },
      "gpuAfterLoad": {
        "usedBytes": 11685330944,
        "freeBytes": 1193279488
      },
      "loadDurationNs": 5303121500,
      "ps": {
        "sizeBytes": 9500916448,
        "sizeVramBytes": 9500916448,
        "hostBytes": 0,
        "contextLength": 8192
      },
      "passes": [
        {
          "warmup": true,
          "evalCount": 64,
          "evalDurationNs": 1448477000,
          "promptEvalCount": 35,
          "promptEvalDurationNs": 134470000
        },
        {
          "warmup": false,
          "evalCount": 64,
          "evalDurationNs": 1479760000,
          "promptEvalCount": 35,
          "promptEvalDurationNs": 22327000
        },
        {
          "warmup": false,
          "evalCount": 64,
          "evalDurationNs": 1516245000,
          "promptEvalCount": 35,
          "promptEvalDurationNs": 25457000
        },
        {
          "warmup": false,
          "evalCount": 64,
          "evalDurationNs": 1535190000,
          "promptEvalCount": 35,
          "promptEvalDurationNs": 24972000
        }
      ],
      "derived": {
        "measuredPasses": 3,
        "generationTokensPerSecond": 42.209537376875105,
        "fullyInVram": true
      }
    },
    {
      "numCtx": 16384,
      "gpuBefore": {
        "usedBytes": 1722810368,
        "freeBytes": 11155800064,
        "utilizationPercent": 26
      },
      "gpuAfterLoad": {
        "usedBytes": 12170821632,
        "freeBytes": 707788800
      },
      "loadDurationNs": 6585788300,
      "ps": {
        "sizeBytes": 11457748007,
        "sizeVramBytes": 10486273802,
        "hostBytes": 971474205,
        "contextLength": 16384
      },
      "passes": [
        {
          "warmup": true,
          "evalCount": 64,
          "evalDurationNs": 1678220000,
          "promptEvalCount": 35,
          "promptEvalDurationNs": 143303000
        },
        {
          "warmup": false,
          "evalCount": 64,
          "evalDurationNs": 1629983000,
          "promptEvalCount": 35,
          "promptEvalDurationNs": 27521000
        },
        {
          "warmup": false,
          "evalCount": 64,
          "evalDurationNs": 1627256000,
          "promptEvalCount": 35,
          "promptEvalDurationNs": 26483000
        },
        {
          "warmup": false,
          "evalCount": 64,
          "evalDurationNs": 1631579000,
          "promptEvalCount": 35,
          "promptEvalDurationNs": 26729000
        }
      ],
      "derived": {
        "measuredPasses": 3,
        "generationTokensPerSecond": 39.26421318504549,
        "fullyInVram": false
      }
    },
    {
      "numCtx": 32768,
      "gpuBefore": {
        "usedBytes": 1447034880,
        "freeBytes": 11431575552,
        "utilizationPercent": 29
      },
      "gpuAfterLoad": {
        "usedBytes": 11848908800,
        "freeBytes": 1029701632
      },
      "loadDurationNs": 6809509800,
      "ps": {
        "sizeBytes": 13987951409,
        "sizeVramBytes": 10187157012,
        "hostBytes": 3800794397,
        "contextLength": 32768
      },
      "passes": [
        {
          "warmup": true,
          "evalCount": 64,
          "evalDurationNs": 2949238000,
          "promptEvalCount": 35,
          "promptEvalDurationNs": 238468000
        },
        {
          "warmup": false,
          "evalCount": 64,
          "evalDurationNs": 2922639000,
          "promptEvalCount": 35,
          "promptEvalDurationNs": 46469000
        },
        {
          "warmup": false,
          "evalCount": 64,
          "evalDurationNs": 2927982000,
          "promptEvalCount": 35,
          "promptEvalDurationNs": 46054000
        },
        {
          "warmup": false,
          "evalCount": 64,
          "evalDurationNs": 2886990000,
          "promptEvalCount": 35,
          "promptEvalDurationNs": 47489000
        }
      ],
      "derived": {
        "measuredPasses": 3,
        "generationTokensPerSecond": 21.898017510886564,
        "fullyInVram": false
      }
    }
  ]
}
