luaLLM/config.example.json at main · ripter/luaLLM · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
{
  "llama_cpp_path": "/usr/local/bin/llama-server",
  "llama_cli_path": "/usr/local/bin/llama-cli",
  "llama_bench_path": "/usr/local/bin/llama-bench",
  "llama_cpp_source_dir": "/home/user/llama.cpp",
  "models_dir": "/home/user/models",
  "recent_models_count": 7,
  "default_port": 8080,
  "default_params": [
    "-c 4096",
    "--host 127.0.0.1",
    "--threads 8"
  ],
  "bench": {
    "default_n": 5,
    "default_warmup": 1,
    "default_ctx": 2048,
    "default_gen": 256,
    "default_batch": 512,
    "default_threads": 8
  },
  "cmake_options": [
    "-DGGML_METAL=ON",
    "-DGGML_METAL_EMBED_LIBRARY=ON",
    "-DGGML_USE_FLASH_ATTENTION=ON",
    "-DLLAMA_BUILD_SERVER=ON",
    "-DLLAMA_BUILD_EXAMPLES=ON",
    "-DCMAKE_BUILD_TYPE=Release"
  ],
  "model_overrides": {
    "codellama": [
      "-c 16384",
      "--gpu-layers 35"
    ],
    "llama%-3.*70b": [
      "-c 8192",
      "--gpu-layers 40",
      "--threads 16"
    ],
    "mistral": [
      "-c 8192"
    ],
    "phi": [
      "-c 4096",
      "--gpu-layers 32"
    ]
  }
}