pezzo

pezzolabsApache-2.0活跃维护

面向开发者的开源 LLMOps 平台,把提示词设计、版本管理、即时下发、团队协作和可观测性串成一条流水线,省去自己拼凑工具的成本。

🕹️ Open-source, developer-first LLMOps platform designed to streamline prompt design, version management, instant delivery, collaboration, troubleshooting, observability and more.

v0.9.2发布于 2024-05-15 12:59 UTC3,280star +1 / 7 天280fork同步于 2026-11-10 17:42 UTC

怎么上手

由 README 提炼,执行前请确认
npm install
docker-compose -f docker-compose.infra.yaml up
npx dotenv-cli -e apps/server/.env -- npx prisma migrate deploy --schema apps/server/prisma/schema.prisma
npx nx serve server
硬件要求按 README 说明

README 未说明,仅要求 Node.js 18+ 与 Docker,依赖服务通过 Docker Compose 启动

有 Docker 镜像
下载:源码 zip·GitHub Releases(安装包)本站不镜像安装包

项目说明

由 README 自动翻译提炼

开源云原生 LLMOps 平台,集中管理提示词、观测 AI 调用并即时发布变更

面向开发者的开源 LLMOps 平台,把提示词设计、版本管理、即时下发、团队协作和可观测性串成一条流水线,省去自己拼凑工具的成本。

核心能力

  • 集中管理提示词设计、版本与协作
  • 观测并排查 AI 调用链路问题
  • 支持缓存以降低成本和延迟
  • 提供 Node.js、Python、LangChain 客户端
  • 基于 PostgreSQL、ClickHouse、Redis 等开源组件

容易踩的坑

需先按 .env.example 创建 .env,使用 Docker 时还需创建 .env.docker;启动服务前必须先执行 Prisma 迁移;开发模式需另开终端运行 npm run graphql:codegen:watch,控制台在 4200 端口、服务在 3000 端口

内容由模型根据 README 生成,可能不完整或不准确,以官方文档为准。看原文 README

最近发版

5 个版本v0.9.2
  1. v0.9.2最新正式版2024-05-15 · 2024-05-15 12:59 UTC

    hotfix: encrypted keys migration

    Full Changelog: https://github.com/pezzolabs/pezzo/compare/v0.9.1...v0.9.2

  2. v0.9.12024-05-14 · 2024-05-14 11:59 UTC
    查看更新说明

    What's Changed

    Full Changelog: https://github.com/pezzolabs/pezzo/compare/v0.9.0...v0.9.1

  3. v0.9.02024-04-28 · 2024-04-28 14:06 UTC
    查看更新说明

    What's Changed

    BREAKING CHANGES 🚨

    In this release, the provider API token encryption method has been changed from AES-CBC to AES-GCM. This introduces a new column in the database, for the ProviderApiKey table. Your users will be required to re-enter their provider API keys and a database migration is required!

    Full Changelog: https://github.com/pezzolabs/pezzo/compare/v0.8.6...v0.9.0

  4. v0.8.62024-04-10 · 2024-04-10 21:21 UTC
    查看更新说明

    Update llm-toolkit for newer OpenAI models

    Full Changelog: https://github.com/pezzolabs/pezzo/compare/v0.8.5...v0.8.6

  5. v0.8.52024-02-03 · 2024-02-03 16:16 UTC
    查看更新说明

    What's Changed

    New Contributors

    Full Changelog: https://github.com/pezzolabs/pezzo/compare/v0.8.4...v0.8.5

用户点评

登录用户的真实使用记录,可按版本和硬件筛选
共 0 条

    还没有人点评。第一个说出真实体验的人最有价值。

    README 原文

    未翻译

    ✨ Features

    Documentation

    Click here to navigate to the Official Pezzo Documentation

    In the documentation, you can find information on how to use Pezzo, its architecture, including tutorials and recipes for varius use cases and LLM providers.

    Supported Clients

    Looking for a client that's not listed here? Open an issue and let us know!

    Getting Started - Docker Compose

    If you simply want to run the full Pezzo stack locally, check out Running With Docker Compose in the documentation.

    If you want to run Pezzo in development mode, continue reading.

    Prerequisites

    Install dependencies

    Install NPM dependencies by running:

    npm install
    

    Set up the environment files

    Pezzo uses a .env file to store environment variables. When using docker, you should also create a .env.docker file.

    See the .env.example file for reference.

    Spin up infrastructure dependencies via Docker Compose

    Pezzo is entirely cloud-native and relies solely on open-source technologies such as PostgreSQL, ClickHouse, Redis and Supertokens.

    You can run these dependencies via Docker Compose:

    docker-compose -f docker-compose.infra.yaml up
    

    Start Pezzo

    Deploy Prisma migrations:

    npx dotenv-cli -e apps/server/.env -- npx prisma migrate deploy --schema apps/server/prisma/schema.prisma
    

    Run the server:

    npx nx serve server
    

    The server is now running. You can verify that by navigating to http://localhost:3000/api/healthz.

    In development mode, you want to run codegen in watch mode, so whenever you make changes to the schema, types are generated automatically. After running the server, run the following in a separate terminal Window:

    npm run graphql:codegen:watch
    

    This will connect codegen directly to the server and keep your GraphQL schema up-to-date as you make changes.

    Finally, you are ready to run the Pezzo Console:

    npx nx serve console
    

    That's it! The Pezzo Console is now accessible at http://localhost:4200 🚀

    Contributing

    We welcome contributions from the community! Please feel free to submit pull requests or create issues for bugs or feature suggestions.

    If you want to contribute but not sure how, join our Discord and we'll be happy to help you out!

    Please check out CONTRIBUTING.md before contributing.

    License

    This repository's source code is available under the Apache 2.0 License.