Asosiy qismga o'tish
AI-Point.uz
← Katalogga qaytish
MCP serverHamjamiyat

LLM Eval Harness

Run a golden set and gate LLM model swaps on accuracy, with latency and cost per request.

Tavsif asl tilda

razilzul · ↓ 0 ta o'rnatish

O'rnatish

Konfiguratsiyani mijoz sozlamalariga qo'shing. <Burchak qavslar> ichidagi qiymatlarni o'zingiznikiga almashtiring.

Konfig fayli: claude_desktop_config.json
{
  "mcpServers": {
    "llm-eval-harness": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "ANTHROPIC_API_KEY",
        "-e",
        "GEMINI_API_KEY",
        "ghcr.io/razilzul/llm-eval-harness:1.0.0"
      ],
      "env": {
        "ANTHROPIC_API_KEY": "<ANTHROPIC_API_KEY>",
        "GEMINI_API_KEY": "<GEMINI_API_KEY>"
      }
    }
  }
}

Yoki bitta buyruq bilan — CLI tanlangan mijoz konfigiga qoʻshadi:

CLI
npx ai-point add io.github.razilzul/llm-eval-harness

Instrumentlar

Instrumentlar ro'yxati serverda sinov muhiti paydo bo'lganda yoki tahririyat uni to'ldirganda ko'rinadi.

Muammo haqida xabar berish ↗ (GitHub yangi oynada ochiladi)