{
  "engine": "llama.cpp",
  "recommended_endpoint_settings": {
    "max_tokens_per_request": 1024,
    "max_concurrent_requests": 2,
    "notes": "Memory scales roughly with (max_concurrent_requests * max_tokens_per_request)."
  },
  "recommended_generation_defaults": {
    "temperature": 1.2,
    "top_p": 0.95,
    "min_p": 0.05,
    "repeat_penalty": 1.08,
    "max_tokens": 2560
  },
  "chat_template": "phi4",
  "gguf_file": "",
  "gguf_quant": "q8_0"
}