{
  "exe": "F:\\Strata\\engine\\strata.exe",
  "args": [
    "--pack",
    "F:\\Strata-data\\packs\\iq3_s",
    "--native",
    "F:\\lmmodels\\Abiray\\Qwen3.8-Flash-Next-GSQ-RCO-IQ3_S\\Qwen3.8-Flash-Next-GSQ-RCO-IQ3_S-00001-of-00002.gguf",
    "--ple-gguf",
    "F:\\lmmodels\\Abiray\\Qwen3.8-Flash-Next-GSQ-RCO-IQ3_S\\Qwen3.8-Flash-Next-GSQ-RCO-IQ3_S-00002-of-00002.gguf",
    "--expert-profile",
    "F:\\Strata\\data\\expert-profile.bin",
    "--expert-cache",
    "auto",
    "--prefill",
    "auto",
    "--spec",
    "4",
    "--spec-min-p",
    "0.5",
    "--mtp",
    "F:\\Strata-data\\mtp\\rt",
    "--max-context",
    "262144",
    "--kv",
    "int8",
    "--kv-resident",
    "32768",
    "--vision",
    "--vram-reserve-mib",
    "1912",
    "--pool-workers",
    "16",
    "--pcie-frac",
    "0.22",
    "--adapt-async",
    "1",
    "--prompt-cache",
    "12",
    "--prompt-cache-every",
    "4096",
    "--conversation-cache-mib",
    "0"
  ],
  "cwd": "F:\\Strata",
  "tokenizer": "F:\\Strata-data\\packs\\iq3_s\\tokenizer",
  "model_name": "qwen3.8-flash-next-iq3_s",
  "log": "F:\\Strata\\strata-iq3_s-single.log",
  "lib_dirs": [
    "F:\\Strata\\.venv\\Lib\\site-packages\\nvidia\\cu13\\bin\\x86_64"
  ],
  "port": 8090,
  "env": {
   "STRATA_ATTN_DAILY": "1",
   "STRATA_EXPDEC": "1",
   "STRATA_B6_MMVQ_ROWS": "1",
   "STRATA_PLE_AUTO_RAM": "0"
  },
  "reasoning_budget_tokens": 512,
  "gpu": 1,
  "vision": {
    "exe": "F:/Strata/engine/vision-on-gpu1.bat",
    "mmproj": "F:\\lmmodels\\Abiray\\Qwen3.8-Flash-Next-GSQ-RCO-IQ3_S\\mmproj-Qwen3.8-Flash-Next-BF16.gguf",
    "model": "F:\\lmmodels\\Abiray\\Qwen3.8-Flash-Next-GSQ-RCO-IQ3_S\\Qwen3.8-Flash-Next-GSQ-RCO-IQ3_S-00001-of-00002.gguf",
    "gpu": true,
    "max_tokens": 1024
  }
}