[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-vespa-engine--vespa":3,"tool-vespa-engine--vespa":64},[4,17,27,35,43,56],{"id":5,"name":6,"github_repo":7,"description_zh":8,"stars":9,"difficulty_score":10,"last_commit_at":11,"category_tags":12,"status":16},3808,"stable-diffusion-webui","AUTOMATIC1111\u002Fstable-diffusion-webui","stable-diffusion-webui 是一个基于 Gradio 构建的网页版操作界面，旨在让用户能够轻松地在本地运行和使用强大的 Stable Diffusion 图像生成模型。它解决了原始模型依赖命令行、操作门槛高且功能分散的痛点，将复杂的 AI 绘图流程整合进一个直观易用的图形化平台。\n\n无论是希望快速上手的普通创作者、需要精细控制画面细节的设计师，还是想要深入探索模型潜力的开发者与研究人员，都能从中获益。其核心亮点在于极高的功能丰富度：不仅支持文生图、图生图、局部重绘（Inpainting）和外绘（Outpainting）等基础模式，还独创了注意力机制调整、提示词矩阵、负向提示词以及“高清修复”等高级功能。此外，它内置了 GFPGAN 和 CodeFormer 等人脸修复工具，支持多种神经网络放大算法，并允许用户通过插件系统无限扩展能力。即使是显存有限的设备，stable-diffusion-webui 也提供了相应的优化选项，让高质量的 AI 艺术创作变得触手可及。",162132,3,"2026-04-05T11:01:52",[13,14,15],"开发框架","图像","Agent","ready",{"id":18,"name":19,"github_repo":20,"description_zh":21,"stars":22,"difficulty_score":23,"last_commit_at":24,"category_tags":25,"status":16},1381,"everything-claude-code","affaan-m\u002Feverything-claude-code","everything-claude-code 是一套专为 AI 编程助手（如 Claude Code、Codex、Cursor 等）打造的高性能优化系统。它不仅仅是一组配置文件，而是一个经过长期实战打磨的完整框架，旨在解决 AI 代理在实际开发中面临的效率低下、记忆丢失、安全隐患及缺乏持续学习能力等核心痛点。\n\n通过引入技能模块化、直觉增强、记忆持久化机制以及内置的安全扫描功能，everything-claude-code 能显著提升 AI 在复杂任务中的表现，帮助开发者构建更稳定、更智能的生产级 AI 代理。其独特的“研究优先”开发理念和针对 Token 消耗的优化策略，使得模型响应更快、成本更低，同时有效防御潜在的攻击向量。\n\n这套工具特别适合软件开发者、AI 研究人员以及希望深度定制 AI 工作流的技术团队使用。无论您是在构建大型代码库，还是需要 AI 协助进行安全审计与自动化测试，everything-claude-code 都能提供强大的底层支持。作为一个曾荣获 Anthropic 黑客大奖的开源项目，它融合了多语言支持与丰富的实战钩子（hooks），让 AI 真正成长为懂上",138956,2,"2026-04-05T11:33:21",[13,15,26],"语言模型",{"id":28,"name":29,"github_repo":30,"description_zh":31,"stars":32,"difficulty_score":23,"last_commit_at":33,"category_tags":34,"status":16},2271,"ComfyUI","Comfy-Org\u002FComfyUI","ComfyUI 是一款功能强大且高度模块化的视觉 AI 引擎，专为设计和执行复杂的 Stable Diffusion 图像生成流程而打造。它摒弃了传统的代码编写模式，采用直观的节点式流程图界面，让用户通过连接不同的功能模块即可构建个性化的生成管线。\n\n这一设计巧妙解决了高级 AI 绘图工作流配置复杂、灵活性不足的痛点。用户无需具备编程背景，也能自由组合模型、调整参数并实时预览效果，轻松实现从基础文生图到多步骤高清修复等各类复杂任务。ComfyUI 拥有极佳的兼容性，不仅支持 Windows、macOS 和 Linux 全平台，还广泛适配 NVIDIA、AMD、Intel 及苹果 Silicon 等多种硬件架构，并率先支持 SDXL、Flux、SD3 等前沿模型。\n\n无论是希望深入探索算法潜力的研究人员和开发者，还是追求极致创作自由度的设计师与资深 AI 绘画爱好者，ComfyUI 都能提供强大的支持。其独特的模块化架构允许社区不断扩展新功能，使其成为当前最灵活、生态最丰富的开源扩散模型工具之一，帮助用户将创意高效转化为现实。",107662,"2026-04-03T11:11:01",[13,14,15],{"id":36,"name":37,"github_repo":38,"description_zh":39,"stars":40,"difficulty_score":23,"last_commit_at":41,"category_tags":42,"status":16},3704,"NextChat","ChatGPTNextWeb\u002FNextChat","NextChat 是一款轻量且极速的 AI 助手，旨在为用户提供流畅、跨平台的大模型交互体验。它完美解决了用户在多设备间切换时难以保持对话连续性，以及面对众多 AI 模型不知如何统一管理的痛点。无论是日常办公、学习辅助还是创意激发，NextChat 都能让用户随时随地通过网页、iOS、Android、Windows、MacOS 或 Linux 端无缝接入智能服务。\n\n这款工具非常适合普通用户、学生、职场人士以及需要私有化部署的企业团队使用。对于开发者而言，它也提供了便捷的自托管方案，支持一键部署到 Vercel 或 Zeabur 等平台。\n\nNextChat 的核心亮点在于其广泛的模型兼容性，原生支持 Claude、DeepSeek、GPT-4 及 Gemini Pro 等主流大模型，让用户在一个界面即可自由切换不同 AI 能力。此外，它还率先支持 MCP（Model Context Protocol）协议，增强了上下文处理能力。针对企业用户，NextChat 提供专业版解决方案，具备品牌定制、细粒度权限控制、内部知识库整合及安全审计等功能，满足公司对数据隐私和个性化管理的高标准要求。",87618,"2026-04-05T07:20:52",[13,26],{"id":44,"name":45,"github_repo":46,"description_zh":47,"stars":48,"difficulty_score":23,"last_commit_at":49,"category_tags":50,"status":16},2268,"ML-For-Beginners","microsoft\u002FML-For-Beginners","ML-For-Beginners 是由微软推出的一套系统化机器学习入门课程，旨在帮助零基础用户轻松掌握经典机器学习知识。这套课程将学习路径规划为 12 周，包含 26 节精炼课程和 52 道配套测验，内容涵盖从基础概念到实际应用的完整流程，有效解决了初学者面对庞大知识体系时无从下手、缺乏结构化指导的痛点。\n\n无论是希望转型的开发者、需要补充算法背景的研究人员，还是对人工智能充满好奇的普通爱好者，都能从中受益。课程不仅提供了清晰的理论讲解，还强调动手实践，让用户在循序渐进中建立扎实的技能基础。其独特的亮点在于强大的多语言支持，通过自动化机制提供了包括简体中文在内的 50 多种语言版本，极大地降低了全球不同背景用户的学习门槛。此外，项目采用开源协作模式，社区活跃且内容持续更新，确保学习者能获取前沿且准确的技术资讯。如果你正寻找一条清晰、友好且专业的机器学习入门之路，ML-For-Beginners 将是理想的起点。",84991,"2026-04-05T10:45:23",[14,51,52,53,15,54,26,13,55],"数据工具","视频","插件","其他","音频",{"id":57,"name":58,"github_repo":59,"description_zh":60,"stars":61,"difficulty_score":10,"last_commit_at":62,"category_tags":63,"status":16},3128,"ragflow","infiniflow\u002Fragflow","RAGFlow 是一款领先的开源检索增强生成（RAG）引擎，旨在为大语言模型构建更精准、可靠的上下文层。它巧妙地将前沿的 RAG 技术与智能体（Agent）能力相结合，不仅支持从各类文档中高效提取知识，还能让模型基于这些知识进行逻辑推理和任务执行。\n\n在大模型应用中，幻觉问题和知识滞后是常见痛点。RAGFlow 通过深度解析复杂文档结构（如表格、图表及混合排版），显著提升了信息检索的准确度，从而有效减少模型“胡编乱造”的现象，确保回答既有据可依又具备时效性。其内置的智能体机制更进一步，使系统不仅能回答问题，还能自主规划步骤解决复杂问题。\n\n这款工具特别适合开发者、企业技术团队以及 AI 研究人员使用。无论是希望快速搭建私有知识库问答系统，还是致力于探索大模型在垂直领域落地的创新者，都能从中受益。RAGFlow 提供了可视化的工作流编排界面和灵活的 API 接口，既降低了非算法背景用户的上手门槛，也满足了专业开发者对系统深度定制的需求。作为基于 Apache 2.0 协议开源的项目，它正成为连接通用大模型与行业专有知识之间的重要桥梁。",77062,"2026-04-04T04:44:48",[15,14,13,26,54],{"id":65,"github_repo":66,"name":67,"description_en":68,"description_zh":69,"ai_summary_zh":69,"readme_en":70,"readme_zh":71,"quickstart_zh":72,"use_case_zh":73,"hero_image_url":74,"owner_login":75,"owner_name":76,"owner_avatar_url":77,"owner_bio":78,"owner_company":79,"owner_location":79,"owner_email":80,"owner_twitter":79,"owner_website":81,"owner_url":82,"languages":83,"stars":122,"forks":123,"last_commit_at":124,"license":125,"difficulty_score":126,"env_os":127,"env_gpu":128,"env_ram":129,"env_deps":130,"category_tags":139,"github_topics":140,"view_count":23,"oss_zip_url":79,"oss_zip_packed_at":79,"status":16,"created_at":154,"updated_at":155,"faqs":156,"releases":185},2057,"vespa-engine\u002Fvespa","vespa","AI + Data, online. https:\u002F\u002Fvespa.ai","Vespa 是一个开源的大数据实时处理平台，专为在海量数据中执行搜索、推荐和个性化应用而设计。它能够在线处理向量、张量、文本及结构化数据，支持在毫秒级延迟内完成数据筛选、机器学习模型推理、结果聚合与返回，即使面对持续更新的海量数据集也能保持高可用性与高性能。\n\n传统方案在处理分布式大规模数据时，往往难以兼顾低延迟与高并发，而 Vespa 通过内置的并行计算引擎解决了这一难题。它允许开发者在数据服务阶段直接嵌入复杂的 AI 模型推理，无需额外的后处理步骤，从而简化了架构并提升了响应速度。\n\nVespa 主要面向后端开发者、数据工程师及算法研究人员，特别适合需要构建高性能搜索系统、实时推荐引擎或大规模数据分析应用的团队。无论是初创公司还是服务于每秒数十万查询的大型互联网服务，都能从中受益。\n\n其技术亮点在于将搜索引擎的能力与机器学习推理深度融合，支持在查询时动态评估模型，并具备弹性伸缩的云原生架构。用户既可以直接部署到云端托管服务，也可以在本地自行搭建和管理集群，灵活适应不同规模的业务需求。","\u003C!-- Copyright Vespa.ai. Licensed under the terms of the Apache 2.0 license. See LICENSE in the project root. -->\n\n\u003Cpicture>\n  \u003Csource media=\"(prefers-color-scheme: dark)\" srcset=\"https:\u002F\u002Fassets.vespa.ai\u002Flogos\u002FVespa-logo-green-RGB.svg\">\n  \u003Csource media=\"(prefers-color-scheme: light)\" srcset=\"https:\u002F\u002Fassets.vespa.ai\u002Flogos\u002FVespa-logo-dark-RGB.svg\">\n  \u003Cimg alt=\"#Vespa\" width=\"200\" src=\"https:\u002F\u002Fassets.vespa.ai\u002Flogos\u002FVespa-logo-dark-RGB.svg\" style=\"margin-bottom: 25px;\">\n\u003C\u002Fpicture>\n\u003Cbr\u002F>\u003Cbr\u002F>\n\n[![Build status](https:\u002F\u002Fbadge.buildkite.com\u002F34f7cb35b91da4f929794c5fd7aa722fc15ca0224ad240270b.svg)](https:\u002F\u002Fbuildkite.com\u002Fvespaai\u002Fvespa-engine-vespa)\n![GitHub License](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flicense\u002Fvespa-engine\u002Fvespa)\n![Maven metadata URL](https:\u002F\u002Fimg.shields.io\u002Fmaven-metadata\u002Fv?metadataUrl=https%3A%2F%2Frepo1.maven.org%2Fmaven2%2Fcom%2Fyahoo%2Fvespa%2Fparent%2Fmaven-metadata.xml)\n\n\n\nSearch, make inferences in and organize vectors, tensors, text and structured data, at serving time and any scale.\n\nThis repository contains all the code required to build and run all of Vespa yourself,\nand where you can see all development as it happens.\nAll the content in this repository is licensed under the Apache 2.0 license.\n\nA new release of Vespa is made from this repository's master branch every morning CET Monday through Thursday.\n\n- Home page: [https:\u002F\u002Fvespa.ai](https:\u002F\u002Fvespa.ai)\n- Documentation: [https:\u002F\u002Fdocs.vespa.ai](https:\u002F\u002Fdocs.vespa.ai)\n- Continuous build: [https:\u002F\u002Ffactory.vespa.ai](https:\u002F\u002Ffactory.vespa.ai)\n- Run applications in the cloud for free: [vespa.ai\u002Ffree-trial](https:\u002F\u002Fvespa.ai\u002Ffree-trial\u002F)\n\n## Table of contents\n\n- [Background](#background)\n- [Install](#install)\n- [Usage](#usage)\n- [Contribute](#contribute)\n- [Building](#building)\n- [License](#license)\n\n## Background\n\nUse cases such as search, recommendation and personalization need to select a subset of data in a large corpus,\nevaluate machine-learned models over the selected data, organize and aggregate it and return it, typically in less\nthan 100 milliseconds, all while the data corpus is continuously changing.\n\nThis is hard to do, especially with large data sets that need to be distributed over multiple nodes and evaluated in\nparallel. Vespa is a platform that performs these operations for you with high availability and performance.\nIt has been in development for many years and is used on several large internet services and apps which serve\nhundreds of thousands of queries from Vespa per second.\n\n## Install\n\nDeploy your Vespa applications to the cloud service: [console.vespa-cloud.com](https:\u002F\u002Fconsole.vespa-cloud.com\u002F),\nor run your own Vespa instance: [https:\u002F\u002Fdocs.vespa.ai\u002Fen\u002Fgetting-started.html](https:\u002F\u002Fdocs.vespa.ai\u002Fen\u002Fgetting-started.html)\n\n## Usage\n\n- The application created in the getting started guides linked above is fully functional and production-ready, but you may want to [add more nodes](https:\u002F\u002Fdocs.vespa.ai\u002Fen\u002Fmultinode-systems.html) for redundancy.\n- See [developing applications](https:\u002F\u002Fdocs.vespa.ai\u002Fen\u002Fdeveloper-guide.html) on adding your own Java components to your Vespa application.\n- [Vespa APIs](https:\u002F\u002Fdocs.vespa.ai\u002Fen\u002Fapi.html) is useful to understand how to interface with Vespa\n- Explore the [sample applications](https:\u002F\u002Fgithub.com\u002Fvespa-engine\u002Fsample-apps\u002Ftree\u002Fmaster)\n- Follow the [Vespa Blog](https:\u002F\u002Fblog.vespa.ai\u002F) for feature updates \u002F use cases\n- Join the [Vespa Slack community](https:\u002F\u002Fslack.vespa.ai\u002F) to ask questions and share feedback\n\nFull documentation is at [https:\u002F\u002Fdocs.vespa.ai](https:\u002F\u002Fdocs.vespa.ai).\n\n## Contribute\n\nWe welcome contributions! See [CONTRIBUTING.md](CONTRIBUTING.md) to learn how to contribute.\n\nIf you want to contribute to the documentation, see\n[https:\u002F\u002Fgithub.com\u002Fvespa-engine\u002Fdocumentation](https:\u002F\u002Fgithub.com\u002Fvespa-engine\u002Fdocumentation)\n\n## Building\n\nYou do not need to build Vespa to use it, but if you want to contribute you need to be able to build the code.\nThis section explains how to build and test Vespa. To understand where to make changes, see [Code-map.md](Code-map.md).\nSome suggested improvements with pointers to code are in [TODO.md](TODO.md).\n\n### Development environment\n\nC++ and Java building is supported on AlmaLinux 8.\nThe Java source can also be built on any platform having Java 17 and Maven 3.8+ installed.\nUse the following guide to set up a complete development environment using Docker\nfor building Vespa, running unit tests and running system tests:\n[Vespa development on AlmaLinux 8](https:\u002F\u002Fgithub.com\u002Fvespa-engine\u002Fdocker-image-dev#vespa-development-on-almalinux-8).\n\n#### Java environment for Mac\n1. Install [JDK17](https:\u002F\u002Fopenjdk.org\u002Fprojects\u002Fjdk\u002F17\u002F), \n   [Maven Version Manager](https:\u002F\u002Fbitbucket.org\u002Fmjensen\u002Fmvnvm\u002Fsrc\u002Fmaster\u002F) and [jEnv](https:\u002F\u002Fwww.jenv.be)\n   through [Homebrew](https:\u002F\u002Fbrew.sh\u002F).\n```sh\nbrew install jenv mvnvm openjdk@17\n```\n\n2. On ARM Macs (M1, M2 etc.), install intel compatibility since [grpc isn't properly maintained](https:\u002F\u002Fgithub.com\u002Fgrpc\u002Fgrpc-java\u002Fissues\u002F7690): \n\n```sh\nsoftwareupdate --install-rosetta\n```\n\n3. For the system Java wrappers to find this JDK, symlink it with\n```sh\nsudo ln -sfn \u002Fopt\u002Fhomebrew\u002Fopt\u002Fopenjdk@17\u002Flibexec\u002Fopenjdk.jdk \u002FLibrary\u002FJava\u002FJavaVirtualMachines\u002Fopenjdk-17.jdk\n```\n\n4. Follow \"Configure your shell\" in https:\u002F\u002Fwww.jenv.be. Configuration is shell specific. For `zsh` use the below commands:\n```sh\necho 'export PATH=\"$HOME\u002F.jenv\u002Fbin:$PATH\"' >> ~\u002F.zshrc\necho 'eval \"$(jenv init -)\"' >> ~\u002F.zshrc\neval \"$(jenv init -)\"\njenv enable-plugin export\nexec $SHELL -l\n```\n\n5. Add JDK17 to jEnv\n```sh\njenv add $(\u002Fusr\u002Flibexec\u002Fjava_home -v 17)\n```\n\n6. Verify configuration with Maven by executing below command in the root of the source code.\n   Output should refer to the JDK and Maven version specified in the [.java-version](.java-version) and [mvnvm.properties](mvnvm.properties).\n```sh\nmvn -v\n```\n\n### Build Java modules\n\n    export MAVEN_OPTS=\"-Xms128m -Xmx1024m\"\n    .\u002Fbootstrap.sh java\n    mvn install --threads 1C\n\nUse this if you only need to build the Java modules, otherwise follow the complete development guide above.\n\n### Run tests for shell scripts (on Mac)\nShell scripts are tested with [BATS](https:\u002F\u002Fbats-core.readthedocs.io\u002Fen\u002Fstable\u002F).\nTo run the tests locally, install the testing framework and its plugins.:\n```bash\nbrew install node\nsudo npm install -g bats bats-assert bats-support bats-mock\n```\nExport the `BATS_PLUGIN_PATH` environment variable to point to the global npm modules directory, which contains the BATS plugins:\n```bash\nexport BATS_PLUGIN_PATH=\"$(npm root -g)\"\n```\nThen run all tests with the following command (from the root of the repository):\n```bash\nbats -r .\n```\nTo run a specific test, use:\n```bash\nbats test_dir\u002Ftest_name.bats\n```\nTests can also be run in IntelliJ IDEA with the [BashSupport Pro](https:\u002F\u002Fplugins.jetbrains.com\u002Fplugin\u002F13841-bashsupport-pro)\nplugin. Ensure the `BATS_PLUGIN_PATH` environment variable is exported before launching the IDE\nto avoid setting it in each run configuration.\n\n## License\n\nCode licensed under the Apache 2.0 license. See [LICENSE](LICENSE) for terms.\n","\u003C!-- 版权归 Vespa.ai 所有。根据 Apache 2.0 许可证条款授权。请参阅项目根目录下的 LICENSE 文件。 -->\n\n\u003Cpicture>\n  \u003Csource media=\"(prefers-color-scheme: dark)\" srcset=\"https:\u002F\u002Fassets.vespa.ai\u002Flogos\u002FVespa-logo-green-RGB.svg\">\n  \u003Csource media=\"(prefers-color-scheme: light)\" srcset=\"https:\u002F\u002Fassets.vespa.ai\u002Flogos\u002FVespa-logo-dark-RGB.svg\">\n  \u003Cimg alt=\"#Vespa\" width=\"200\" src=\"https:\u002F\u002Fassets.vespa.ai\u002Flogos\u002FVespa-logo-dark-RGB.svg\" style=\"margin-bottom: 25px;\">\n\u003C\u002Fpicture>\n\u003Cbr\u002F>\u003Cbr\u002F>\n\n[![构建状态](https:\u002F\u002Fbadge.buildkite.com\u002F34f7cb35b91da4f929794c5fd7aa722fc15ca0224ad240270b.svg)](https:\u002F\u002Fbuildkite.com\u002Fvespaai\u002Fvespa-engine-vespa)\n![GitHub 许可证](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flicense\u002Fvespa-engine\u002Fvespa)\n![Maven 元数据 URL](https:\u002F\u002Fimg.shields.io\u002Fmaven-metadata\u002Fv?metadataUrl=https%3A%2F%2Frepo1.maven.org%2Fmaven2%2Fcom%2Fyahoo%2Fvespa%2Fparent%2Fmaven-metadata.xml)\n\n\n\n在服务时以任意规模对向量、张量、文本和结构化数据进行搜索、推理和组织。\n\n本仓库包含构建和运行 Vespa 所需的所有代码，并且您可以实时查看所有开发进展。\n本仓库中的所有内容均采用 Apache 2.0 许可证授权。\n\n每周一至周四上午中欧时间，都会从该仓库的 master 分支发布 Vespa 的新版本。\n\n- 首页：[https:\u002F\u002Fvespa.ai](https:\u002F\u002Fvespa.ai)\n- 文档：[https:\u002F\u002Fdocs.vespa.ai](https:\u002F\u002Fdocs.vespa.ai)\n- 持续集成：[https:\u002F\u002Ffactory.vespa.ai](https:\u002F\u002Ffactory.vespa.ai)\n- 在云端免费运行应用：[vespa.ai\u002Ffree-trial](https:\u002F\u002Fvespa.ai\u002Ffree-trial\u002F)\n\n## 目录\n\n- [背景](#background)\n- [安装](#install)\n- [使用](#usage)\n- [贡献](#contribute)\n- [构建](#building)\n- [许可证](#license)\n\n## 背景\n\n诸如搜索、推荐和个性化等用例需要从庞大的语料库中筛选出一部分数据，\n对这些数据运行机器学习模型进行评估，再对其进行组织和聚合后返回结果，整个过程通常需要在不到 100 毫秒内完成，\n而且数据语料库还在不断变化。\n\n要做到这一点非常困难，尤其是在处理需要分布到多个节点并并行计算的大规模数据集时。Vespa 是一个能够为您高效、高可用地执行这些操作的平台。\n它经过多年的发展，已被多家大型互联网服务和应用所采用，每秒可处理数十万次来自 Vespa 的查询请求。\n\n## 安装\n\n将您的 Vespa 应用部署到云服务：[console.vespa-cloud.com](https:\u002F\u002Fconsole.vespa-cloud.com\u002F)，\n或自行运行一个 Vespa 实例：[https:\u002F\u002Fdocs.vespa.ai\u002Fen\u002Fgetting-started.html](https:\u002F\u002Fdocs.vespa.ai\u002Fen\u002Fgetting-started.html)\n\n## 使用\n\n- 上文链接的入门指南中创建的应用程序已经完全可用且可用于生产环境，但您可能希望[添加更多节点](https:\u002F\u002Fdocs.vespa.ai\u002Fen\u002Fmultinode-systems.html)以提高冗余性。\n- 有关如何为您的 Vespa 应用程序添加自定义 Java 组件，请参阅[开发应用程序](https:\u002F\u002Fdocs.vespa.ai\u002Fen\u002Fdeveloper-guide.html)。\n- [Vespa API](https:\u002F\u002Fdocs.vespa.ai\u002Fen\u002Fapi.html) 有助于理解如何与 Vespa 进行交互。\n- 浏览[示例应用程序](https:\u002F\u002Fgithub.com\u002Fvespa-engine\u002Fsample-apps\u002Ftree\u002Fmaster)。\n- 关注[Vespa 博客](https:\u002F\u002Fblog.vespa.ai\u002F)以获取功能更新和用例信息。\n- 加入[Vespa Slack 社区](https:\u002F\u002Fslack.vespa.ai\u002F)提问并分享反馈。\n\n完整文档请访问[https:\u002F\u002Fdocs.vespa.ai](https:\u002F\u002Fdocs.vespa.ai)。\n\n## 贡献\n\n我们欢迎各方贡献！请参阅[CONTRIBUTING.md](CONTRIBUTING.md)，了解如何参与贡献。\n如果您想为文档做出贡献，请访问\n[https:\u002F\u002Fgithub.com\u002Fvespa-engine\u002Fdocumentation](https:\u002F\u002Fgithub.com\u002Fvespa-engine\u002Fdocumentation)。\n\n## 构建\n\n您无需构建 Vespa 即可使用它，但如果您想参与贡献，则需要能够编译代码。\n本节将说明如何构建和测试 Vespa。要了解应在何处进行修改，请参阅[Code-map.md](Code-map.md)。\n一些带有代码指针的改进建议可在[TODO.md](TODO.md)中找到。\n\n### 开发环境\n\nC++ 和 Java 的编译支持 AlmaLinux 8 系统。\nJava 源代码也可以在任何已安装 Java 17 和 Maven 3.8+ 的平台上编译。\n请参考以下指南，使用 Docker 设置完整的开发环境，用于构建 Vespa、运行单元测试和系统测试：\n[Vespa 在 AlmaLinux 8 上的开发](https:\u002F\u002Fgithub.com\u002Fvespa-engine\u002Fdocker-image-dev#vespa-development-on-almalinux-8)。\n\n#### Mac 上的 Java 环境\n1. 通过 Homebrew 安装 [JDK17](https:\u002F\u002Fopenjdk.org\u002Fprojects\u002Fjdk\u002F17\u002F)、\n   [Maven 版本管理器](https:\u002F\u002Fbitbucket.org\u002Fmjensen\u002Fmvnvm\u002Fsrc\u002Fmaster\u002F) 和 [jEnv](https:\u002F\u002Fwww.jenv.be)：\n```sh\nbrew install jenv mvnvm openjdk@17\n```\n\n2. 对于 ARM 架构的 Mac（如 M1、M2 等），由于 [grpc 尚未得到良好维护](https:\u002F\u002Fgithub.com\u002Fgrpc\u002Fgrpc-java\u002Fissues\u002F7690)，请安装 Intel 兼容层：\n\n```sh\nsoftwareupdate --install-rosetta\n```\n\n3. 为了让系统 Java 包装器找到此 JDK，需创建符号链接：\n```sh\nsudo ln -sfn \u002Fopt\u002Fhomebrew\u002Fopt\u002Fopenjdk@17\u002Flibexec\u002Fopenjdk.jdk \u002FLibrary\u002FJava\u002FJavaVirtualMachines\u002Fopenjdk-17.jdk\n```\n\n4. 按照 https:\u002F\u002Fwww.jenv.be 中的“配置您的 shell”步骤操作。配置因 shell 而异。对于 `zsh`，请使用以下命令：\n```sh\necho 'export PATH=\"$HOME\u002F.jenv\u002Fbin:$PATH\"' >> ~\u002F.zshrc\necho 'eval \"$(jenv init -)\"' >> ~\u002F.zshrc\neval \"$(jenv init -)\"\njenv enable-plugin export\nexec $SHELL -l\n```\n\n5. 将 JDK17 添加到 jEnv：\n```sh\njenv add $(\u002Fusr\u002Flibexec\u002Fjava_home -v 17)\n```\n\n6. 通过在源代码根目录下执行以下命令来验证配置是否正确。输出应显示 [.java-version](.java-version) 和 [mvnvm.properties](mvnvm.properties) 中指定的 JDK 和 Maven 版本。\n```sh\nmvn -v\n```\n\n### 编译 Java 模块\n\n    export MAVEN_OPTS=\"-Xms128m -Xmx1024m\"\n    .\u002Fbootstrap.sh java\n    mvn install --threads 1C\n\n如果您只需要编译 Java 模块，可以使用此方法；否则，请按照上述完整的开发指南操作。\n\n### 在 Mac 上运行 Shell 脚本测试\nShell 脚本使用 [BATS](https:\u002F\u002Fbats-core.readthedocs.io\u002Fen\u002Fstable\u002F) 进行测试。\n要在本地运行测试，请先安装测试框架及其插件：\n```bash\nbrew install node\nsudo npm install -g bats bats-assert bats-support bats-mock\n```\n导出 `BATS_PLUGIN_PATH` 环境变量，使其指向全局 npm 模块目录，该目录包含 BATS 插件：\n```bash\nexport BATS_PLUGIN_PATH=\"$(npm root -g)\"\n```\n然后在仓库根目录下运行以下命令以执行所有测试：\n```bash\nbats -r .\n```\n要运行特定的测试，可以使用：\n```bash\nbats test_dir\u002Ftest_name.bats\n```\n此外，也可以在 IntelliJ IDEA 中使用 [BashSupport Pro](https:\u002F\u002Fplugins.jetbrains.com\u002Fplugin\u002F13841-bashsupport-pro) 插件来运行测试。请确保在启动 IDE 之前已导出 `BATS_PLUGIN_PATH` 环境变量，以免需要在每个运行配置中单独设置。\n\n## 许可证\n\n代码采用 Apache 2.0 许可证授权。详细条款请参阅 [LICENSE](LICENSE) 文件。","# Vespa 快速上手指南\n\nVespa 是一个高性能的大数据处理与实时计算引擎，适用于搜索、推荐、个性化及向量检索等场景。它能够在毫秒级延迟下，对海量数据进行过滤、机器学习模型推理、聚合与排序。\n\n## 环境准备\n\n### 系统要求\n*   **推荐系统**: AlmaLinux 8 (官方构建环境)。\n*   **通用支持**: 任何安装了 Java 17 和 Maven 3.8+ 的操作系统（Linux, macOS, Windows via WSL）。\n*   **macOS 特别说明**:\n    *   需安装 Rosetta 2 (针对 M1\u002FM2 等 ARM 芯片)，因为部分 gRPC 组件依赖 Intel 架构兼容层。\n    *   推荐使用 Docker 进行开发以获得一致的环境体验。\n\n### 前置依赖\n请确保已安装以下工具：\n*   **JDK**: OpenJDK 17\n*   **构建工具**: Maven 3.8+\n*   **容器化 (可选但推荐)**: Docker\n\n**macOS 用户快速配置命令:**\n```bash\n# 1. 安装必要工具 (通过 Homebrew)\nbrew install jenv mvnvm openjdk@17\n\n# 2. ARM 芯片 (M1\u002FM2) 安装 Intel 兼容层\nsoftwareupdate --install-rosetta\n\n# 3. 创建 JDK 软链接以便系统识别\nsudo ln -sfn \u002Fopt\u002Fhomebrew\u002Fopt\u002Fopenjdk@17\u002Flibexec\u002Fopenjdk.jdk \u002FLibrary\u002FJava\u002FJavaVirtualMachines\u002Fopenjdk-17.jdk\n\n# 4. 配置 jEnv (以 zsh 为例)\necho 'export PATH=\"$HOME\u002F.jenv\u002Fbin:$PATH\"' >> ~\u002F.zshrc\necho 'eval \"$(jenv init -)\"' >> ~\u002F.zshrc\neval \"$(jenv init -)\"\njenv enable-plugin export\nexec $SHELL -l\n\n# 5. 将 JDK 17 添加到 jEnv 管理\njenv add $(\u002Fusr\u002Flibexec\u002Fjava_home -v 17)\n\n# 6. 验证版本\nmvn -v\n```\n\n## 安装步骤\n\nVespa 提供两种主要使用方式：**云端托管**（最快上手）和 **本地自建**。\n\n### 方案 A：使用 Vespa Cloud (推荐新手)\n无需配置本地环境，直接在云端免费试用。\n1. 访问 [Vespa Cloud Console](https:\u002F\u002Fconsole.vespa-cloud.com\u002F)。\n2. 注册账号并创建新应用。\n3. 按照网页指引部署即可。\n\n### 方案 B：本地自建 (Docker 方式)\n适合需要在本地调试或私有化部署的开发者。\n\n1. **拉取并运行 Vespa Docker 镜像**:\n   ```bash\n   docker run --detach --name vespa --hostname vespa-container \\\n     -p 8080:8080 -p 19071:19071 \\\n     vespaengine\u002Fvespa\n   ```\n   *注：启动可能需要几分钟，可通过 `docker logs --follow vespa` 查看进度，直到出现 \"Vespa started\" 字样。*\n\n2. **验证安装**:\n   访问 `http:\u002F\u002Flocalhost:8080\u002FApplicationStatus` 确认服务状态为 `up`。\n\n3. **部署示例应用**:\n   确保本地已安装 `vespa-cli` (或通过 Docker 执行)，部署一个简单的搜索应用：\n   ```bash\n   # 如果使用本地 vespa-cli\n   vespa deploy --wait-for-deployment\n   \n   # 或者直接使用官方提供的 sample-apps (需先 git clone)\n   # cd sample-apps\u002Fbasic-search\n   # vespa deploy --wait-for-deployment\n   ```\n\n## 基本使用\n\n以下演示如何向 Vespa 写入数据并执行查询。假设你已部署了基础的 `basic-search` 示例应用。\n\n### 1. 写入文档 (Feed)\n创建一个名为 `feed.json` 的文件，包含测试数据：\n```json\n[\n  {\n    \"put\": \"id:music:music::1\",\n    \"fields\": {\n      \"title\": \"Best of Blues\",\n      \"artist\": \"B.B. King\",\n      \"year\": 2000,\n      \"category_scores\": {\n        \"cells\": [\n          {\"address\": {\"cat\": \"pop\"}, \"value\": 0.1},\n          {\"address\": {\"cat\": \"rock\"}, \"value\": 0.2},\n          {\"address\": {\"cat\": \"blues\"}, \"value\": 0.9}\n        ]\n      }\n    }\n  },\n  {\n    \"put\": \"id:music:music::2\",\n    \"fields\": {\n      \"title\": \"Greatest Hits\",\n      \"artist\": \"Queen\",\n      \"year\": 1981,\n      \"category_scores\": {\n        \"cells\": [\n          {\"address\": {\"cat\": \"pop\"}, \"value\": 0.8},\n          {\"address\": {\"cat\": \"rock\"}, \"value\": 0.9},\n          {\"address\": {\"cat\": \"blues\"}, \"value\": 0.1}\n        ]\n      }\n    }\n  }\n]\n```\n\n执行写入命令：\n```bash\nvespa feed feed.json\n```\n\n### 2. 执行查询 (Query)\n使用 `vespa query` 或 `curl` 进行搜索。例如，查找艺术家为 \"Queen\" 的歌曲：\n\n```bash\nvespa query \"select * from music where artist contains 'Queen'\"\n```\n\n或者使用原生 HTTP API：\n```bash\ncurl -H \"Content-Type: application\u002Fjson\" \\\n  \"http:\u002F\u002Flocalhost:8080\u002Fsearch\u002F?query=queen&hits=2\"\n```\n\n### 3. 下一步\n*   **自定义开发**: 参考 [开发者指南](https:\u002F\u002Fdocs.vespa.ai\u002Fen\u002Fdeveloper-guide.html) 添加自定义 Java 组件。\n*   **更多示例**: 浏览 [sample-apps](https:\u002F\u002Fgithub.com\u002Fvespa-engine\u002Fsample-apps) 仓库获取搜索、推荐、向量检索等完整案例。\n*   **社区交流**: 加入 [Vespa Slack](https:\u002F\u002Fslack.vespa.ai\u002F) 获取技术支持。","某大型跨境电商平台需要构建一个实时个性化商品推荐系统，要求在用户浏览瞬间，从亿级商品库中结合文本语义、图像特征向量及库存价格等结构化数据，完成毫秒级的混合检索与排序。\n\n### 没有 vespa 时\n- **架构复杂且延迟高**：开发团队需维护 Elasticsearch 做文本搜索、Faiss 做向量检索、Redis 存属性数据，多次服务调用导致整体响应时间远超 100 毫秒。\n- **实时性差**：商品库存变化或价格调整无法即时同步到检索引擎，用户常看到已下架或价格错误的商品推荐。\n- **模型落地困难**：难以在服务端直接运行机器学习模型对召回结果进行重排序，必须将数据导出到外部计算集群，流程冗长。\n- **运维成本高昂**：多套异构系统的数据一致性保障和分布式扩容极其复杂，大促期间频繁出现性能瓶颈。\n\n### 使用 vespa 后\n- **统一引擎极速响应**：vespa 单引擎同时支持文本、向量和结构化数据的混合查询，将原本分散的链路整合，稳定实现 50 毫秒内的端到端返回。\n- **数据实时更新**：利用 vespa 的流式写入能力，商品状态变更秒级可见，确保推荐结果永远基于最新库存和价格。\n- **原生模型推理**：直接在 vespa 节点上部署并执行 ONNX 格式的深度学习模型，在检索过程中即时完成个性化打分与排序。\n- **弹性伸缩简便**：面对流量洪峰，只需简单增加节点即可线性提升吞吐量，无需关心底层数据分片与同步逻辑。\n\nvespa 通过“数据 +AI\"的一体化架构，让海量数据的实时智能决策变得像普通数据库查询一样简单高效。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fvespa-engine_vespa_6611a6ea.png","vespa-engine","Vespa","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Fvespa-engine_127d11eb.png","Vespa is an open-source platform for applications that need low-latency computation over large structured, text and vector data.",null,"info@vespa.ai","https:\u002F\u002Fvespa.ai","https:\u002F\u002Fgithub.com\u002Fvespa-engine",[84,88,92,96,100,104,108,112,115,118],{"name":85,"color":86,"percentage":87},"Java","#b07219",51.9,{"name":89,"color":90,"percentage":91},"C++","#f34b7d",44.7,{"name":93,"color":94,"percentage":95},"Go","#00ADD8",1.4,{"name":97,"color":98,"percentage":99},"CMake","#DA3434",0.8,{"name":101,"color":102,"percentage":103},"Shell","#89e051",0.4,{"name":105,"color":106,"percentage":107},"Python","#3572A5",0.2,{"name":109,"color":110,"percentage":111},"JavaScript","#f1e05a",0.1,{"name":113,"color":114,"percentage":111},"HTML","#e34c26",{"name":116,"color":117,"percentage":111},"C","#555555",{"name":119,"color":120,"percentage":121},"TLA","#4b0079",0,6856,706,"2026-04-05T09:18:44","Apache-2.0",4,"Linux (AlmaLinux 8), macOS","未说明","未说明 (构建 Java 模块时建议 Maven 堆内存至少 1GB)",{"notes":131,"python":128,"dependencies":132},"1. C++ 和 Java 的完整构建环境官方支持 AlmaLinux 8，推荐使用提供的 Docker 镜像进行开发、单元测试和系统测试。\n2. Java 源码可在任何安装了 Java 17 和 Maven 3.8+ 的平台上构建。\n3. macOS (特别是 M1\u002FM2 等 ARM 架构) 用户需安装 Rosetta 以兼容 grpc，并需手动配置 JDK  symlink 以便系统识别。\n4. 若仅需构建 Java 模块，可直接使用 Maven，无需完整 C++ 环境。\n5. Shell 脚本测试依赖 BATS 框架及其插件。",[133,134,135,136,137,138],"Java 17","Maven 3.8+","C++ 编译器 (AlmaLinux 8)","Docker (推荐用于开发环境)","Node.js (用于 Shell 脚本测试)","BATS (用于 Shell 脚本测试)",[14,54,13,51,15],[67,141,142,143,144,145,146,147,148,149,150,151,152,153],"search-engine","big-data","ai","serving-recommendation","machine-learning","server","java","vector-search","rag","search","vector","tensor","vector-database","2026-03-27T02:49:30.150509","2026-04-06T05:27:29.159252",[157,162,167,172,177,181],{"id":158,"question_zh":159,"answer_zh":160,"source_url":161},9374,"如何配置多节点 Config Server 集群？","需要在所有节点的 $VESPA_HOME\u002Fconf\u002Fvespa\u002Fdefault-env.txt 文件中设置环境变量。关键配置包括：\n1. 设置 VESPA_CONFIGSERVERS 为所有 config server 节点的主机名列表。\n2. 确保 hosts.xml 文件中正确定义了所有主机及其别名。\n3. 注意：虽然可以将 admin\u002Fconfig 角色与 container\u002Fcontent 角色部署在同一节点上，但在某些配置下（特别是 Docker 环境外）可能会遇到错误。如果遇到问题，尝试将 admin\u002Fconfig 节点独立出来部署。\n示例 env 配置：\noverride VESPA_CONFIGSERVERS vespa-1,vespa-2,vespa-3","https:\u002F\u002Fgithub.com\u002Fvespa-engine\u002Fvespa\u002Fissues\u002F5419",{"id":163,"question_zh":164,"answer_zh":165,"source_url":166},9375,"如何通过 API 动态更新或删除 Schema（模式文件）而无需重启？","可以使用 Vespa 的应用部署 API 进行增量更新。具体步骤如下：\n1. 创建新会话：\ncurl -X POST \"http:\u002F\u002Flocalhost:19071\u002Fapplication\u002Fv2\u002Ftenant\u002Fdefault\u002Fsession?from=...\"\n2. 获取返回的会话 ID (例如 VESPA_SESSION=4)。\n3. 上传更新的 services.xml 和新的 schema 文件，或删除旧的 schema 文件：\ncurl -X PUT --data-binary @services.xml \"...\u002Fsession\u002F${VESPA_SESSION}\u002Fcontent\u002Fservices.xml\"\ncurl -X DELETE \"...\u002Fsession\u002F${VESPA_SESSION}\u002Fcontent\u002Fschemas\u002Fold_schema.sd\"\n4. 准备并激活部署：\ncurl -X PUT \"...\u002Fsession\u002F${VESPA_SESSION}\u002Fprepared?applicationName=default\"\n此方法允许在不重启服务的情况下动态调整字段定义。","https:\u002F\u002Fgithub.com\u002Fvespa-engine\u002Fvespa\u002Fissues\u002F26104",{"id":168,"question_zh":169,"answer_zh":170,"source_url":171},9376,"Vespa Proton 启动时占用大量内存（即使没有文档数据）是正常的吗？","是的，这在一定程度上是正常的。Vespa 会预分配内存用于永久数据结构。对于拥有多个 schema（例如 39 个）的情况，启动时占用较多内存（如 20GB）可能是由于多值属性（multivalue attributes）的静态内存开销。\n建议：\n1. 升级到 Vespa 8.181.15 或更高版本，该版本将多值属性的静态内存使用量减少了高达 40%。\n2. 注意监控日志中的 SizeClass 分布，确认内存分配是否符合预期。\n3. 实际写入文档后，内存占比通常会趋于合理范围。","https:\u002F\u002Fgithub.com\u002Fvespa-engine\u002Fvespa\u002Fissues\u002F26350",{"id":173,"question_zh":174,"answer_zh":175,"source_url":176},9377,"Vespa 是否支持基于索引（Index）的模糊匹配（Fuzzy Match），以避免 Attribute 匹配的高内存消耗？","Vespa 的原生索引模型主要支持 text\u002Fword\u002Fexact\u002Fgram 模式，不直接像 Lucene 那样支持 term 级别的 fuzzy match。但是，可以通过自定义 Searcher 来实现类似效果：\n1. 在 services.xml 中替换默认的 NGramSearcher，将其逻辑从 AND 改为 OR，以扩大召回范围。\n配置示例：\n\u003Csearch>\n  \u003Cchain id='default' inherits='vespa'\u002F>\n  \u003Cprovider cluster='mycontent' excludes='com.yahoo.search.querytransform.NGramSearcher' id='local' type='local'>\n    \u003Csearcher bundle='basic-application' id='ai.vespa.example.OrNGramSearcher'\u002F>\n  \u003C\u002Fprovider>\n\u003C\u002Fsearch>\n2. 编写 Java 代码继承 NGramSearcher 并重写相关逻辑（参考 album-recommendation-java 示例应用）。\n注意：改用 OR 逻辑会显著增加查询命中的数量，需权衡性能。","https:\u002F\u002Fgithub.com\u002Fvespa-engine\u002Fvespa\u002Fissues\u002F9371",{"id":178,"question_zh":179,"answer_zh":180,"source_url":176},9378,"如何在 Vespa 应用中添加自定义 Java 代码（如自定义 Searcher）？","需要遵循标准的 Maven 项目结构：\n1. 在 Vespa 应用项目的 src\u002Fmain\u002Fjava 目录下放置 Java 源代码。\n2. 确保项目包含正确的 pom.xml 配置以构建 bundle。\n3. 参考官方文档 searcher-development 或查看 sample-apps 仓库中的 'album-recommendation-java' 示例应用，了解完整的目录结构和依赖配置。\n4. 编译后将生成的 jar 包部署到 Vespa 应用中，并在 services.xml 中引用该 searcher。",{"id":182,"question_zh":183,"answer_zh":184,"source_url":161},9379,"是否可以在同一节点上运行多种 Vespa 角色（如同时作为 Config 节点和内容节点）？","可以。Vespa 允许在同一台机器上运行多个角色（例如同时作为 admin\u002Fconfig 节点和 container\u002Fcontent 节点）。\n但在实际操作中发现：\n1. 在某些配置下（特别是非 Docker 环境或特定 hosts.xml 配置），混合部署可能会导致启动错误。\n2. 如果遇到无法复现的错误，建议尝试将 admin\u002Fconfig 角色分离到独立节点（例如新增一个 vespa-config 节点）。\n3. 在 Docker 环境中通常没有强制分离的必要，但在生产环境大规模部署时，分离角色有助于资源隔离和稳定性。",[186,191,195,199,203,207,211,215,219,223,227,231,235,240,244,249,253,257,261,265],{"id":187,"version":188,"summary_zh":189,"released_at":190},106725,"v8.667.16","\u003C!-- Copyright Vespa.ai. Licensed under the terms of the Apache 2.0 license. See LICENSE in the project root. -->\n\nThe command-line tool for Vespa.ai.\n\nUse it on Vespa instances running locally, remotely or in the cloud.\nPrefer web service API's to this in production.\n\nSee [Vespa documentation](https:\u002F\u002Fdocs.vespa.ai) and [getting started with Vespa\nCLI](https:\u002F\u002Fdocs.vespa.ai\u002Fen\u002Freference\u002Fclients\u002Fvespa-cli.html).\n\nRun `make` to build and test - make sure to use go 1.18 or higher.\n","2026-03-31T09:05:17",{"id":192,"version":193,"summary_zh":189,"released_at":194},106726,"v8.665.18","2026-03-30T05:24:25",{"id":196,"version":197,"summary_zh":189,"released_at":198},106727,"v8.660.21","2026-03-19T03:31:18",{"id":200,"version":201,"summary_zh":189,"released_at":202},106728,"v8.658.52","2026-03-17T03:36:11",{"id":204,"version":205,"summary_zh":189,"released_at":206},106729,"v8.653.22","2026-03-09T03:22:25",{"id":208,"version":209,"summary_zh":189,"released_at":210},106730,"v8.651.80","2026-03-05T07:49:43",{"id":212,"version":213,"summary_zh":189,"released_at":214},106731,"v8.650.16","2026-03-02T05:24:13",{"id":216,"version":217,"summary_zh":189,"released_at":218},106732,"v8.648.10","2026-02-25T05:05:56",{"id":220,"version":221,"summary_zh":189,"released_at":222},106733,"v8.646.22","2026-02-20T13:18:41",{"id":224,"version":225,"summary_zh":189,"released_at":226},106734,"v8.640.27","2026-02-11T09:08:00",{"id":228,"version":229,"summary_zh":189,"released_at":230},106735,"v8.639.59","2026-02-10T03:25:12",{"id":232,"version":233,"summary_zh":189,"released_at":234},106736,"v8.638.30","2026-02-09T03:23:40",{"id":236,"version":237,"summary_zh":238,"released_at":239},106737,"lsp-v2.4.9","The Language-server for Vespa schemas\n\nUse the jar file to integrate the language server into your favorite editor.\n\nFor Visual Studio Code and IntelliJ the language server should also be available in the marketplace for the editor.\n\n# Schema Language Server in Neovim\n\n## Requirements\nRequires Java 17 or newer.\n\nOptional: [lspconfig](https:\u002F\u002Fgithub.com\u002Fneovim\u002Fnvim-lspconfig) plugin for nvim.\n\n## Installation\nDownload `schema-language-server-jar-with-dependencies.jar`.\n\n### Using lspconfig\nThe language server is registered at `lspconfig` as `vespa_ls`. If you have `lspconfig` installed, all that needs to \nbe done is to enable the language server.\n\nRegister `.sd`, `.profile` and `.yql` as filetypes (in `init.lua`):\n```lua\nvim.filetype.add {\n  extension = {\n    profile = 'sd',\n    sd = 'sd',\n    yql = 'yql'\n  }\n}\n```\n\nCreate a config for schema language server (in `init.lua`):\n```lua\nvim.lsp.config('vespa_ls', {\n    cmd = { 'java', '-jar', '\u002Fpath\u002Fto\u002Fvespa-language-server_X.X.X.jar' },\n    -- on_attach = ...\n})\n\nvim.lsp.enable('vespa_ls')\n```\n\n### Manual Installation\nIf you don't want to use lspconfig you can refer to the [LSP documentation for Neovim](https:\u002F\u002Fneovim.io\u002Fdoc\u002Fuser\u002Flsp.html) for manually registering the server.\n## What's New\n- The 'tensorFromStructs' rank feature.\n- Fix bugs related to the 'foreach' rank feature.\n- Publish to the Open VSX registry.\n- General stability improvements.\n","2026-02-05T14:00:37",{"id":241,"version":242,"summary_zh":189,"released_at":243},106738,"v8.636.27","2026-02-04T03:22:25",{"id":245,"version":246,"summary_zh":247,"released_at":248},106739,"lsp-v2.4.8","The Language-server for Vespa schemas\n\nUse the jar file to integrate the language server into your favorite editor.\n\nFor Visual Studio Code and IntelliJ the language server should also be available in the marketplace for the editor.\n\n# Schema Language Server in Neovim\n\n## Requirements\nRequires Java 17 or newer.\n\nOptional: [lspconfig](https:\u002F\u002Fgithub.com\u002Fneovim\u002Fnvim-lspconfig) plugin for nvim.\n\n## Installation\nDownload `schema-language-server-jar-with-dependencies.jar`.\n\n### Using lspconfig\nThe language server is registered at `lspconfig` as `vespa_ls`. If you have `lspconfig` installed, all that needs to \nbe done is to enable the language server.\n\nRegister `.sd`, `.profile` and `.yql` as filetypes (in `init.lua`):\n```lua\nvim.filetype.add {\n  extension = {\n    profile = 'sd',\n    sd = 'sd',\n    yql = 'yql'\n  }\n}\n```\n\nCreate a config for schema language server (in `init.lua`):\n```lua\nvim.lsp.config('vespa_ls', {\n    cmd = { 'java', '-jar', '\u002Fpath\u002Fto\u002Fvespa-language-server_X.X.X.jar' },\n    -- on_attach = ...\n})\n\nvim.lsp.enable('vespa_ls')\n```\n\n### Manual Installation\nIf you don't want to use lspconfig you can refer to the [LSP documentation for Neovim](https:\u002F\u002Fneovim.io\u002Fdoc\u002Fuser\u002Flsp.html) for manually registering the server.\n## What's New\n- Support for multiple inheritance of match-features.\n- Fixed formatting bugs.\n","2026-01-27T09:39:11",{"id":250,"version":251,"summary_zh":189,"released_at":252},106740,"v8.631.39","2026-01-22T03:27:10",{"id":254,"version":255,"summary_zh":189,"released_at":256},106741,"v8.629.20","2026-01-15T15:38:43",{"id":258,"version":259,"summary_zh":189,"released_at":260},106742,"v8.624.72","2026-01-06T14:36:36",{"id":262,"version":263,"summary_zh":189,"released_at":264},106743,"v8.620.35","2025-12-11T11:37:57",{"id":266,"version":267,"summary_zh":189,"released_at":268},106744,"v8.618.24","2025-12-08T03:21:52"]