{
  "schemaVersion": "1.0",
  "updates": [],
  "companyNews": [],
  "id": "b703",
  "language": "zh",
  "name": "Ragas",
  "originalName": "Ragas",
  "library": "tools",
  "category": "模型开发",
  "categoryKey": "模型开发",
  "categoryKeys": [
    "模型开发"
  ],
  "categories": [
    "模型开发"
  ],
  "entryPoints": [],
  "summary": "用传统指标和模型评分评估 RAG 与语言模型应用，并从资料或运行数据构建测试样本。",
  "introduction": "Ragas 提供评估语言模型应用的指标、测试数据生成和反馈流程。开发者把输入、回答及任务所需上下文组织为样例，选择传统或基于模型的评分方法，比较 RAG 等系统的输出。也可从已有资料或运行数据生成更贴近目标场景的测试集。评价器本身的模型、提示和样本分布会影响分数，适合用于持续比较而非绝对准确性认证。",
  "capabilities": [],
  "company": {
    "name": "VibrantLabs",
    "originalName": "VibrantLabs"
  },
  "officialUrl": "https://docs.ragas.io/en/stable/",
  "profileUpdatedOn": "2026-10-22",
  "sources": [
    {
      "kind": "product",
      "url": "https://docs.ragas.io/en/stable/"
    }
  ],
  "news": [],
  "links": {
    "interactive": "products/ragas--b703.html",
    "reading": "read/zh/products/ragas--b703.html",
    "markdown": "read/zh/products/ragas--b703.md",
    "json": "read/zh/products/ragas--b703.json"
  }
}