[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-HamaWhiteGG--langchain-java":3,"tool-HamaWhiteGG--langchain-java":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":70,"readme_en":71,"readme_zh":72,"quickstart_zh":73,"use_case_zh":74,"hero_image_url":75,"owner_login":76,"owner_name":77,"owner_avatar_url":78,"owner_bio":79,"owner_company":79,"owner_location":80,"owner_email":81,"owner_twitter":79,"owner_website":79,"owner_url":82,"languages":83,"stars":88,"forks":89,"last_commit_at":90,"license":91,"difficulty_score":23,"env_os":92,"env_gpu":93,"env_ram":93,"env_deps":94,"category_tags":98,"github_topics":99,"view_count":23,"oss_zip_url":79,"oss_zip_packed_at":79,"status":16,"created_at":107,"updated_at":108,"faqs":109,"releases":147},2021,"HamaWhiteGG\u002Flangchain-java","langchain-java","Java version of LangChain, while empowering LLM for Big Data.","LangChain Java 是 LangChain 的 Java 版本，专为 Java 开发者打造，帮助他们在大数据生态中轻松集成大语言模型（LLM）能力。它提供了一套简洁的 API，支持 SQL 查询、RAG（检索增强生成）、摘要生成、API 调用等常见 LLM 应用场景，并原生兼容 Spark 和 Flink 等大数据框架，让 Java 工程师无需切换语言即可构建智能数据应用。  \n\n传统上，LLM 应用多依赖 Python 生态，而 LangChain Java 填补了 Java 项目在 AI 能力接入上的空白，尤其适合企业级大数据平台中已有 Java 技术栈的团队。它支持 OpenAI、ChatGLM、Ollama 等主流模型，以及 Pinecone、Milvus 等向量数据库，开箱即用。开发者只需引入 Maven 依赖，配置 API 密钥，即可快速搭建基于 LLM 的问答系统、智能分析代理或自动化数据处理流程。  \n\n适合熟悉 Java 和大数据技术的开发者与数据工程师使用，尤其在金融、电信、电商等依赖 Java 后端系统的行业中有较高实用价值。其对 Spark SQL 和","LangChain Java 是 LangChain 的 Java 版本，专为 Java 开发者打造，帮助他们在大数据生态中轻松集成大语言模型（LLM）能力。它提供了一套简洁的 API，支持 SQL 查询、RAG（检索增强生成）、摘要生成、API 调用等常见 LLM 应用场景，并原生兼容 Spark 和 Flink 等大数据框架，让 Java 工程师无需切换语言即可构建智能数据应用。  \n\n传统上，LLM 应用多依赖 Python 生态，而 LangChain Java 填补了 Java 项目在 AI 能力接入上的空白，尤其适合企业级大数据平台中已有 Java 技术栈的团队。它支持 OpenAI、ChatGLM、Ollama 等主流模型，以及 Pinecone、Milvus 等向量数据库，开箱即用。开发者只需引入 Maven 依赖，配置 API 密钥，即可快速搭建基于 LLM 的问答系统、智能分析代理或自动化数据处理流程。  \n\n适合熟悉 Java 和大数据技术的开发者与数据工程师使用，尤其在金融、电信、电商等依赖 Java 后端系统的行业中有较高实用价值。其对 Spark SQL 和 Flink SQL 的原生代理支持，是区别于其他 LLM 框架的独特亮点。","# 🦜️ LangChain Java\n\nJava version of LangChain, while empowering LLM for BigData. \n\nIt serves as a bridge to the realm of LLM within the Big Data domain, primarily in the Java stack.\n![Introduction to Langchain-Java.png](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FHamaWhiteGG_langchain-java_readme_bc2355dc1d6b.png)\n\n> If you are interested, you can add me on WeChat: HamaWhite, or send email to [me](mailto:baisongxx@gmail.com).\n\n## 1. What is this?\n\nThis is the Java language implementation of LangChain, which makes it as easy as possible to develop LLM-powered applications.\n![Langchain overview.png](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FHamaWhiteGG_langchain-java_readme_fcc7616be50d.png)\n\nThe following example in the [langchain-example](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples).\n\n- [SQL Chain](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchains\u002FSqlChainExample.java)\n- [API Chain](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchains\u002FApiChainExample.java)\n- [RAG Milvus](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchains\u002FMilvusExample.java)\n- [RAG Pinecone](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchains\u002FRetrievalQaExample.java)\n- [Summarization](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchains\u002FSummarizationExample.java)\n- [Google Search Agent](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fagents\u002FChatAgentExample.java)\n- [Spark SQL Agent](langchain-bigdata\u002Flangchain-spark\u002Fsrc\u002Ftest\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fagents\u002Ftoolkits\u002Fspark\u002Fsql\u002Ftoolkit\u002FSparkSqlToolkitTest.java)\n- [Flink SQL Agent](langchain-bigdata\u002Flangchain-flink\u002Fsrc\u002Ftest\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fagents\u002Ftoolkits\u002Fflink\u002Fsql\u002Ftoolkit\u002FFlinkSqlToolkitTest.java)\n\n## 2. Integrations\n\n### 2.1 LLMs\n- [OpenAI](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fllms\u002FOpenAIExample.java) [[stream](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fllms\u002FStreamOpenAIExample.java)]\n- [Azure OpenAI](openai-client\u002Fsrc\u002Ftest\u002Fjava\u002Fcom\u002Fhw\u002Fopenai\u002FAzureOpenAiClientTest.java)\n- [ChatGLM2](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fllms\u002FChatGLMExample.java)\n- [Ollama](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fllms\u002FOllamaExample.java)\n\n### 2.2 Vector stores\n- [Pinecone](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fvectorstores\u002FPineconeExample.java)\n- [Milvus](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchains\u002FMilvusExample.java)\n\n## 3. Quickstart Guide\nThe API documentation is available at the following link:   \n[https:\u002F\u002Fhamawhitegg.github.io\u002Flangchain-java](https:\u002F\u002Fhamawhitegg.github.io\u002Flangchain-java)\n\n### 3.1 Maven Repository\nPrerequisites for building:\n* Java 17 or later\n* Unix-like environment (we use Linux, Mac OS X)\n* Maven (we recommend version 3.8.6 and require at least 3.5.4)\n\n [![Maven Central](https:\u002F\u002Fimg.shields.io\u002Fmaven-central\u002Fv\u002Fio.github.hamawhitegg\u002Flangchain-core)](https:\u002F\u002Fmaven-badges.herokuapp.com\u002Fmaven-central\u002Fio.github.hamawhitegg\u002Flangchain-core)\n```xml\n\u003Cdependency>\n    \u003CgroupId>io.github.hamawhitegg\u003C\u002FgroupId>\n    \u003CartifactId>langchain-core\u003C\u002FartifactId>\n    \u003Cversion>0.2.1\u003C\u002Fversion>\n\u003C\u002Fdependency>\n```\n\n### 3.2 Environment Setup\nUsing LangChain will usually require integrations with one or more model providers, data stores, apis, etc. \nFor this example, we will be using OpenAI’s APIs.\n\nWe will then need to set the environment variable.\n```shell\nexport OPENAI_API_KEY=xxx\n\n# If a proxy is needed, set the OPENAI_PROXY environment variable.\nexport OPENAI_PROXY=http:\u002F\u002Fhost:port\n```\n\nIf you want to set the API key and proxy dynamically, you can use the openaiApiKey and openaiProxy parameter when initiating OpenAI class.\n```java\nvar llm = OpenAI.builder()\n        .openaiOrganization(\"xxx\")\n        .openaiApiKey(\"xxx\")\n        .openaiProxy(\"http:\u002F\u002Fhost:port\")\n        .requestTimeout(16)\n        .build()\n        .init();\n```\n\n### 3.3 LLMs\nGet predictions from a language model. The basic building block of LangChain is the LLM, which takes in text and generates more text.\n\n[OpenAI Example](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fllms\u002FOpenAIExample.java)\n```java\nvar llm = OpenAI.builder()\n        .temperature(0.9f)\n        .build()\n        .init();\n\nvar result = llm.predict(\"What would be a good company name for a company that makes colorful socks?\");\nprint(result);\n```\nAnd now we can pass in text and get predictions!\n```shell\nFeetful of Fun\n```\n### 3.4 Chat models\n\nChat models are a variation on language models. While chat models use language models under the hood, the interface they expose is a bit different: rather than expose a \"text in, text out\" API, they expose an interface where \"chat messages\" are the inputs and outputs.\n\n[OpenAI Chat Example](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchat\u002Fmodels\u002FChatExample.java)\n```java\nvar chat = ChatOpenAI.builder()\n        .temperature(0)\n        .build()\n        .init();\n\nvar result = chat.predictMessages(List.of(new HumanMessage(\"Translate this sentence from English to French. I love programming.\")));\nprintln(result);\n```\n\n```shell\nAIMessage{content='J'adore la programmation.', additionalKwargs={}}\n```\n\nIt is useful to understand how chat models are different from a normal LLM, but it can often be handy to just be able to treat them the same. LangChain makes that easy by also exposing an interface through which you can interact with a chat model as you would a normal LLM. You can access this through the `predict` interface.\n```java\nvar output = chat.predict(\"Translate this sentence from English to French. I love programming.\");\nprintln(output);\n```\n```shell\nJ'adore la programmation.\n```\n\n### 3.5 Chains\n\nNow that we've got a model and a prompt template, we'll want to combine the two. Chains give us a way to link (or chain) together multiple primitives, like models, prompts, and other chains.\n\n#### 3.5.1 LLMs\nThe simplest and most common type of chain is an LLMChain, which passes an input first to a PromptTemplate and then to an LLM. We can construct an LLM chain from our existing model and prompt template.\n\n[LLM Chain Example](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchains\u002FLlmChainExample.java)\n```java\nvar prompt = PromptTemplate.fromTemplate(\"What is a good name for a company that makes {product}?\");\n\nvar chain = new LLMChain(llm, prompt);\nvar result = chain.run(\"colorful socks\");\nprintln(result);\n```\n```shell\nFeetful of Fun\n```\n#### 3.5.2 Chat models\nThe `LLMChain` can be used with chat models as well:\n\n[LLM Chat Chain Example](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchains\u002FChatChainExample.java)\n```java\nvar template = \"You are a helpful assistant that translates {input_language} to {output_language}.\";\nvar systemMessagePrompt = SystemMessagePromptTemplate.fromTemplate(template);\nvar humanMessagePrompt = HumanMessagePromptTemplate.fromTemplate(\"{text}\");\nvar chatPrompt = ChatPromptTemplate.fromMessages(List.of(systemMessagePrompt, humanMessagePrompt));\n\nvar chain = new LLMChain(chat, chatPrompt);\nvar result = chain.run(Map.of(\"input_language\", \"English\", \"output_language\", \"French\", \"text\", \"I love programming.\"));\nprintln(result);\n```\n```shell\nJ'adore la programmation.\n```\n\n#### 3.5.1 SQL Chains Example\nLLMs make it possible to interact with SQL databases using natural language, and LangChain offers SQL Chains to build and run SQL queries based on natural language prompts.\n\n![SQL chains.png](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FHamaWhiteGG_langchain-java_readme_ffc5a2db2af8.png)\n\n[SQL Chain Example](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchains\u002FSqlChainExample.java)\n```java\nvar database = SQLDatabase.fromUri(\"jdbc:mysql:\u002F\u002F127.0.0.1:3306\u002Fdemo\", \"xxx\", \"xxx\");\n\nvar chain = SQLDatabaseChain.fromLLM(llm, database);\nvar result = chain.run(\"How many students are there?\");\nprintln(result);\n\nresult = chain.run(\"Who got zero score? Show me her parent's contact information.\");\nprintln(result);\n```\n```shell\nThere are 6 students.\n\nThe parent of the student who got zero score is Tracy and their contact information is 088124.\n```\n\nAvailable Languages are as follows.\n\n| Language           | Value |\n|--------------------|-------|\n| English(default)   | en_US |\n| Portuguese(Brazil) | pt_BR |\n\nIf you want to choose other language instead english, just set environment variable on your host. If you not set, then **en-US** will be default\n```shell\nexport USE_LANGUAGE=pt_BR\n```\n\n### 3.6 Agents\nOur first chain ran a pre-determined sequence of steps. To handle complex workflows, we need to be able to dynamically choose actions based on inputs.\n\nAgents do just this: they use a language model to determine which actions to take and in what order. Agents are given access to tools, and they repeatedly choose a tool, run the tool, and observe the output until they come up with a final answer.\n\nSet the appropriate environment variables.\n```shell\nexport SERPAPI_API_KEY=xxx\n```\n\n#### 3.6.1 Google Search Agent Example\nTo augment OpenAI's knowledge beyond 2021 and computational abilities through the use of the Search and Calculator tools.\n![Google agent example.png](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FHamaWhiteGG_langchain-java_readme_ee3dc6567219.png)\n\n[Google Search Agent Example](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fagents\u002FChatAgentExample.java)\n```java\n\u002F\u002F the 'llm-math' tool uses an LLM\nvar tools = loadTools(List.of(\"serpapi\", \"llm-math\"), llm);\n\nvar agent = initializeAgent(tools, chat, AgentType.CHAT_ZERO_SHOT_REACT_DESCRIPTION);\nvar query = \"How many countries and regions participated in the 2023 Hangzhou Asian Games?\" +\n        \"What is that number raised to the .023 power?\";\n\nagent.run(query);\n```\n![Google agent example output.png](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FHamaWhiteGG_langchain-java_readme_ab2262a4666e.png)\n\n## 4. Run Test Cases from Source\n\n```shell\ngit clone https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java.git\ncd langchain-java\n\n# export JAVA_HOME=JDK17_INSTALL_HOME && mvn clean test\nmvn clean test\n```\n\nThis project uses Spotless to format the code. If you make any modifications, please remember to format the code using the following command.\n\n```shell\n# export JAVA_HOME=JDK17_INSTALL_HOME && mvn spotless:apply\nmvn spotless:apply\n```\n\n## 5. Support\nDon’t hesitate to ask!\n\n[Open an issue](https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fissues) if you find a bug in langchain-java.\n\n## 6. Reward\nIf the project has been helpful to you, you can treat me to a cup of coffee.\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FHamaWhiteGG_langchain-java_readme_a515a743c248.png\" alt=\"Appreciation code\" style=\"width:40%;\">\n> This is a WeChat appreciation code.","# 🦜️ LangChain Java\n\nLangChain 的 Java 版本，助力大数据领域的大型语言模型（LLM）应用。\n\n它充当了大数据领域中 LLM 领域的桥梁，主要面向 Java 技术栈。\n![Langchain-Java 简介.png](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FHamaWhiteGG_langchain-java_readme_bc2355dc1d6b.png)\n\n> 如果您感兴趣，可以加我微信：HamaWhite，或发送邮件至 [我](mailto:baisongxx@gmail.com)。\n\n## 1. 这是什么？\n\n这是 LangChain 的 Java 语言实现，旨在让开发基于 LLM 的应用尽可能简单。\n![Langchain 概览.png](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FHamaWhiteGG_langchain-java_readme_fcc7616be50d.png)\n\n以下示例位于 [langchain-example](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples) 中。\n\n- [SQL 链](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchains\u002FSqlChainExample.java)\n- [API 链](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchains\u002FApiChainExample.java)\n- [RAG Milvus](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchains\u002FMilvusExample.java)\n- [RAG Pinecone](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchains\u002FRetrievalQaExample.java)\n- [摘要生成](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchains\u002FSummarizationExample.java)\n- [Google 搜索代理](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fagents\u002FChatAgentExample.java)\n- [Spark SQL 代理](langchain-bigdata\u002Flangchain-spark\u002Fsrc\u002Ftest\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fagents\u002Ftoolkits\u002Fspark\u002Fsql\u002Ftoolkit\u002FSparkSqlToolkitTest.java)\n- [Flink SQL 代理](langchain-bigdata\u002Flangchain-flink\u002Fsrc\u002Ftest\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fagents\u002Ftoolkits\u002Fflink\u002Fsql\u002Ftoolkit\u002FFlinkSqlToolkitTest.java)\n\n## 2. 集成\n\n### 2.1 大型语言模型\n- [OpenAI](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fllms\u002FOpenAIExample.java) [[流式处理](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fllms\u002FStreamOpenAIExample.java)]\n- [Azure OpenAI](openai-client\u002Fsrc\u002Ftest\u002Fjava\u002Fcom\u002Fhw\u002Fopenai\u002FAzureOpenAiClientTest.java)\n- [ChatGLM2](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fllms\u002FChatGLMExample.java)\n- [Ollama](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fllms\u002FOllamaExample.java)\n\n### 2.2 向量存储\n- [Pinecone](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fvectorstores\u002FPineconeExample.java)\n- [Milvus](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchains\u002FMilvusExample.java)\n\n## 3. 快速入门指南\nAPI 文档可在以下链接获取：  \n[https:\u002F\u002Fhamawhitegg.github.io\u002Flangchain-java](https:\u002F\u002Fhamawhitegg.github.io\u002Flangchain-java)\n\n### 3.1 Maven 仓库\n构建前提条件：\n* Java 17 或更高版本\n* 类 Unix 环境（我们使用 Linux、Mac OS X）\n* Maven（推荐版本 3.8.6，最低要求 3.5.4）\n\n[![Maven Central](https:\u002F\u002Fimg.shields.io\u002Fmaven-central\u002Fv\u002Fio.github.hamawhitegg\u002Flangchain-core)](https:\u002F\u002Fmaven-badges.herokuapp.com\u002Fmaven-central\u002Fio.github.hamawhitegg\u002Flangchain-core)\n```xml\n\u003Cdependency>\n    \u003CgroupId>io.github.hamawhitegg\u003C\u002FgroupId>\n    \u003CartifactId>langchain-core\u003C\u002FartifactId>\n    \u003Cversion>0.2.1\u003C\u002Fversion>\n\u003C\u002Fdependency>\n```\n\n### 3.2 环境设置\n使用 LangChain 通常需要与一个或多个模型提供商、数据存储、API 等集成。\n在本示例中，我们将使用 OpenAI 的 API。\n\n接下来我们需要设置环境变量。\n```shell\nexport OPENAI_API_KEY=xxx\n\n# 如果需要代理，设置 OPENAI_PROXY 环境变量。\nexport OPENAI_PROXY=http:\u002F\u002Fhost:port\n```\n\n如果您想动态设置 API 密钥和代理，可以在初始化 OpenAI 类时使用 openaiApiKey 和 openaiProxy 参数。\n```java\nvar llm = OpenAI.builder()\n        .openaiOrganization(\"xxx\")\n        .openaiApiKey(\"xxx\")\n        .openaiProxy(\"http:\u002F\u002Fhost:port\")\n        .requestTimeout(16)\n        .build()\n        .init();\n```\n\n### 3.3 大型语言模型\n从语言模型获取预测结果。LangChain 的基本构建块是 LLM，它接收文本并生成更多文本。\n\n[OpenAI 示例](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fllms\u002FOpenAIExample.java)\n```java\nvar llm = OpenAI.builder()\n        .temperature(0.9f)\n        .build()\n        .init();\n\nvar result = llm.predict(\"一家生产彩色袜子的公司，起个好听的公司名会是什么？\");\nprint(result);\n```\n现在我们可以传入文本并获取预测结果！\n```shell\n趣味十足\n```\n### 3.4 聊天模型\n\n聊天模型是语言模型的一种变体。虽然聊天模型底层使用的是语言模型，但它们提供的接口略有不同：不是“输入文本，输出文本”的 API，而是以“聊天消息”作为输入和输出的接口。\n\n[OpenAI 聊天示例](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchat\u002Fmodels\u002FChatExample.java)\n```java\nvar chat = ChatOpenAI.builder()\n        .temperature(0)\n        .build()\n        .init();\n\nvar result = chat.predictMessages(List.of(new HumanMessage(\"将这句话从英语翻译成法语。我喜欢编程。\")));\nprintln(result);\n```\n\n```shell\nAIMessage{content='J'adore la programmation.', additionalKwargs={}}\n```\n\n了解聊天模型与普通 LLM 的区别很有用，但很多时候直接把它们当作普通 LLM 来处理也很方便。LangChain 通过提供一个接口，让您能像操作普通 LLM 一样与聊天模型交互。您可以通过 `predict` 接口访问这一功能。\n```java\nvar output = chat.predict(\"将这句话从英语翻译成法语。我喜欢编程。\");\nprintln(output);\n```\n```shell\nJ'adore la programmation.\n```\n\n### 3.5 链\n\n现在我们已经有了一个模型和一个提示模板，接下来就需要将两者结合起来。链为我们提供了一种将多个原语（如模型、提示和其他链）链接（或串联）起来的方式。\n\n#### 3.5.1 大语言模型\n最简单也最常见的链类型是LLMChain，它会先将输入传递给PromptTemplate，然后再传递给大语言模型。我们可以用现有的模型和提示模板来构建一个LLMChain。\n\n[LLM Chain 示例](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchains\u002FLlmChainExample.java)\n```java\nvar prompt = PromptTemplate.fromTemplate(\"What is a good name for a company that makes {product}?\");\n\nvar chain = new LLMChain(llm, prompt);\nvar result = chain.run(\"colorful socks\");\nprintln(result);\n```\n```shell\nFeetful of Fun\n```\n#### 3.5.2 聊天模型\n`LLMChain`同样可以用于聊天模型：\n\n[LLM 聊天链 示例](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchains\u002FChatChainExample.java)\n```java\nvar template = \"You are a helpful assistant that translates {input_language} to {output_language}.\";\nvar systemMessagePrompt = SystemMessagePromptTemplate.fromTemplate(template);\nvar humanMessagePrompt = HumanMessagePromptTemplate.fromTemplate(\"{text}\");\nvar chatPrompt = ChatPromptTemplate.fromMessages(List.of(systemMessagePrompt, humanMessagePrompt));\n\nvar chain = new LLMChain(chat, chatPrompt);\nvar result = chain.run(Map.of(\"input_language\", \"English\", \"output_language\", \"French\", \"text\", \"I love programming.\"));\nprintln(result);\n```\n```shell\nJ'adore la programmation.\n```\n\n#### 3.5.1 SQL 链 示例\n大语言模型使得我们能够通过自然语言与SQL数据库交互，LangChain提供了SQL链，以便根据自然语言提示构建并运行SQL查询。\n\n![SQL 链.png](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FHamaWhiteGG_langchain-java_readme_ffc5a2db2af8.png)\n\n[SQL 链 示例](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchains\u002FSqlChainExample.java)\n```java\nvar database = SQLDatabase.fromUri(\"jdbc:mysql:\u002F\u002F127.0.0.1:3306\u002Fdemo\", \"xxx\", \"xxx\");\n\nvar chain = SQLDatabaseChain.fromLLM(llm, database);\nvar result = chain.run(\"How many students are there?\");\nprintln(result);\n\nresult = chain.run(\"Who got zero score? Show me her parent's contact information.\");\nprintln(result);\n```\n```shell\nThere are 6 students.\n\nThe parent of the student who got zero score is Tracy and their contact information is 088124.\n```\n\n可用的语言如下。\n\n| 语言           | 值 |\n|--------------------|-------|\n| 英语（默认）   | en_US |\n| 葡萄牙语（巴西） | pt_BR |\n\n如果你想选择其他语言而非英语，只需在你的主机上设置环境变量。如果没有设置，则默认为**en-US**。\n```shell\nexport USE_LANGUAGE=pt_BR\n```\n\n### 3.6 智能体\n我们第一个链执行的是预定义的步骤序列。为了处理复杂的工作流，我们需要能够根据输入动态地选择行动。\n\n智能体正是这样做的：它们使用语言模型来决定采取哪些行动以及以什么顺序执行。智能体被赋予工具访问权限，它们会反复选择一个工具、运行该工具，并观察输出，直到得出最终答案。\n\n设置合适的环境变量。\n```shell\nexport SERPAPI_API_KEY=xxx\n```\n\n#### 3.6.1 Google 搜索智能体示例\n通过使用搜索和计算器工具，增强OpenAI超越2021年的知识储备和计算能力。\n![Google 智能体示例.png](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FHamaWhiteGG_langchain-java_readme_ee3dc6567219.png)\n\n[Google 搜索智能体示例](langchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fagents\u002FChatAgentExample.java)\n```java\n\u002F\u002F 'llm-math' 工具使用大语言模型\nvar tools = loadTools(List.of(\"serpapi\", \"llm-math\"), llm);\n\nvar agent = initializeAgent(tools, chat, AgentType.CHAT_ZERO_SHOT_REACT_DESCRIPTION);\nvar query = \"How many countries and regions participated in the 2023 Hangzhou Asian Games?\" +\n        \"What is that number raised to the .023 power?\";\n\nagent.run(query);\n```\n![Google 智能体示例输出.png](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FHamaWhiteGG_langchain-java_readme_ab2262a4666e.png)\n\n## 4. 从源码运行测试用例\n\n```shell\ngit clone https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java.git\ncd langchain-java\n\n# export JAVA_HOME=JDK17_INSTALL_HOME && mvn clean test\nmvn clean test\n```\n\n本项目使用Spotless来格式化代码。如果你做了任何修改，请记得使用以下命令格式化代码。\n\n```shell\n# export JAVA_HOME=JDK17_INSTALL_HOME && mvn spotless:apply\nmvn spotless:apply\n```\n\n## 5. 支持\n如有任何问题，请随时提问！\n\n如果在langchain-java中发现了bug，请[打开一个问题](https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fissues)。\n\n## 6. 打赏\n如果这个项目对你有所帮助，你可以请我喝杯咖啡。\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FHamaWhiteGG_langchain-java_readme_a515a743c248.png\" alt=\"赞赏码\" style=\"width:40%;\">\n> 这是一个微信赞赏码。","# LangChain Java 快速上手指南\n\n## 环境准备\n\n- **Java 版本**：JDK 17 或更高  \n- **操作系统**：Linux \u002F macOS（推荐 Unix-like 环境）  \n- **构建工具**：Maven 3.5.4+（推荐 3.8.6）  \n\n> 如需使用 OpenAI 等外部服务，请确保网络可访问，或配置代理。\n\n## 安装步骤\n\n1. 添加 Maven 依赖（中央仓库）：\n\n```xml\n\u003Cdependency>\n    \u003CgroupId>io.github.hamawhitegg\u003C\u002FgroupId>\n    \u003CartifactId>langchain-core\u003C\u002FartifactId>\n    \u003Cversion>0.2.1\u003C\u002Fversion>\n\u003C\u002Fdependency>\n```\n\n2. 设置 OpenAI API 密钥（如使用 OpenAI）：\n\n```shell\nexport OPENAI_API_KEY=xxx\n# 可选：设置代理\nexport OPENAI_PROXY=http:\u002F\u002Fhost:port\n```\n\n3. 克隆示例项目（可选）：\n\n```shell\ngit clone https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java.git\ncd langchain-java\nmvn clean test\n```\n\n> 如需代码格式化：`mvn spotless:apply`\n\n## 基本使用\n\n### 最简 LLM 调用示例\n\n```java\nimport com.hw.langchain.llms.OpenAI;\n\nvar llm = OpenAI.builder()\n    .temperature(0.9f)\n    .build()\n    .init();\n\nvar result = llm.predict(\"What would be a good company name for a company that makes colorful socks?\");\nSystem.out.println(result);\n```\n\n**输出示例**：\n```\nFeetful of Fun\n```\n\n### Chat 模型调用示例\n\n```java\nimport com.hw.langchain.chat.models.ChatOpenAI;\nimport com.hw.langchain.messages.HumanMessage;\n\nvar chat = ChatOpenAI.builder()\n    .temperature(0)\n    .build()\n    .init();\n\nvar result = chat.predict(\"Translate this sentence from English to French. I love programming.\");\nSystem.out.println(result);\n```\n\n**输出示例**：\n```\nJ'adore la programmation.\n```\n\n> 所有示例代码详见：[langchain-examples](https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Ftree\u002Fdev\u002Flangchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples)","某大型银行的数据分析团队每天需要从数百万条客户交易记录中自动提取异常行为模式，并生成可读的风控报告。团队基于Java技术栈构建大数据平台，使用Spark和Flink处理海量数据，但缺乏高效方式将LLM能力融入现有流程，人工撰写报告耗时且易遗漏关键线索。\n\n### 没有 langchain-java 时\n- 团队需手动编写复杂SQL查询来筛选可疑交易，每次调整规则都需重新部署代码，响应慢。\n- 生成的原始数据报表全是数字和代码，业务人员无法直接理解，需额外人工解释。\n- 想引入LLM做自然语言摘要或问答，但缺乏Java生态的成熟库，只能依赖外部Python服务，增加系统耦合和延迟。\n- 向量检索（如Milvus）用于相似交易匹配时，需自己封装API调用和嵌入逻辑，开发成本高。\n- 无法动态调用外部API（如征信系统）辅助判断，自动化程度低，依赖人工复核。\n\n### 使用 langchain-java 后\n- 通过`Spark SQL Agent`直接在Spark作业中调用LLM，自动将SQL查询结果转化为自然语言风险摘要，无需人工干预。\n- 利用`Summarization`链对每日异常交易列表自动生成结构化风控简报，业务部门可直接阅读。\n- 集成`Milvus`向量库，实现“相似欺诈模式”自动检索，仅用几行代码即可完成嵌入与匹配，替代原手动规则库。\n- 借助`API Chain`动态调用内部征信接口，结合LLM推理判断交易风险等级，实现端到端自动化决策。\n- 所有逻辑均在Java应用内完成，无需跨语言调用，系统延迟降低70%，运维复杂度大幅下降。\n\nlangchain-java 让Java大数据团队无需跳出技术栈，就能无缝接入大模型能力，把枯燥的数据处理变成智能决策引擎。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FHamaWhiteGG_langchain-java_176fa08d.png","HamaWhiteGG","HamaWhite","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002FHamaWhiteGG_1d4b8aa1.jpg",null,"Hangzhou","baisongxx@gmail.com","https:\u002F\u002Fgithub.com\u002FHamaWhiteGG",[84],{"name":85,"color":86,"percentage":87},"Java","#b07219",100,568,107,"2026-04-03T19:06:24","Apache-2.0","Linux, macOS","未说明",{"notes":95,"python":93,"dependencies":96},"项目基于 Java 17+，需配置 Maven 3.5.4+；使用 OpenAI 等外部服务需设置 API 密钥和代理；支持连接 MySQL、Pinecone、Milvus 等数据源；代码格式化需使用 mvn spotless:apply；部分功能依赖外部 API（如 SerpAPI、OpenAI），需网络访问。",[97],"io.github.hamawhitegg:langchain-core:0.2.1",[13,26],[100,101,102,103,104,105,106],"langchain","llm","java","large-language-models","openai","langchian-java","sql-chain","2026-03-27T02:49:30.150509","2026-04-06T06:54:42.334320",[110,115,120,125,130,135,139,143],{"id":111,"question_zh":112,"answer_zh":113,"source_url":114},9162,"是否支持本地部署的向量存储，例如 ChromaDB？","目前支持 Milvus 作为本地向量存储，示例代码见：[MilvusExample](https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fblob\u002Fmain\u002Flangchain-examples\u002Fsrc\u002Fmain\u002Fjava\u002Fcom\u002Fhw\u002Flangchain\u002Fexamples\u002Fchains\u002FMilvusExample.java)。可通过 ConnectParam 配置本地主机和端口，结合 OpenAIEmbeddings 使用。ChromaDB 尚未原生支持，但可参考 Milvus 实现方式自行扩展。","https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fissues\u002F38",{"id":116,"question_zh":117,"answer_zh":118,"source_url":119},9163,"项目为何要求 JDK 17？能否在 JDK 8 上运行？","项目默认使用 JDK 17，但已创建独立的 jdk8 分支支持 Java 8。用户可直接使用 [jdk8 分支](https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Ftree\u002Fjdk8)。有用户通过 GPT 将 JDK 17 代码自动转换为 JDK 8 兼容代码，重点处理了字符串语法和新特性，可作为临时解决方案。","https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fissues\u002F65",{"id":121,"question_zh":122,"answer_zh":123,"source_url":124},9164,"向 Pinecone 插入数据时出现 HTTP 400 错误，如何解决？","HTTP 400 错误通常是由于向量维度不匹配导致。Pinecone 的 text-embedding-ada-002 模型要求向量维度为 1536。请检查创建索引时的 dimension 参数是否设置为 1536，例如：CreateIndexRequest.builder().name(INDEX_NAME).dimension(1536).build()。","https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fissues\u002F73",{"id":126,"question_zh":127,"answer_zh":128,"source_url":129},9165,"使用 ChatGLM2 时开启 withHistory(true) 报错，如何正确处理对话历史？","Java 版本的 ChatGLM2 实现中，history 应为 List\u003CList\u003CString>> 格式（即 [[问题, 回答], ...]），而 API 期望的是 List\u003CTuple\u003CString, String>>。建议将历史对话封装为嵌套列表结构传入，避免直接拼接 prompt。维护者指出，ConversationChain 已在内部管理历史，推荐使用其内存机制而非直接传递原始 history 参数。","https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fissues\u002F125",{"id":131,"question_zh":132,"answer_zh":133,"source_url":134},9166,"调用 Pinecone 时出现 HTTP 500 错误，可能原因是什么？","HTTP 500 错误通常由服务端异常引起，可能是索引未就绪或请求格式错误。请确保在创建索引后调用 awaitIndexReady(client) 等待索引完成初始化，再执行 upsert 或 describeIndexStats 操作。同时检查 API 密钥、环境变量（pineconeEnv）是否正确配置。","https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fissues\u002F79",{"id":136,"question_zh":137,"answer_zh":138,"source_url":124},9167,"如何在 Java 中正确配置 Pinecone 客户端以避免连接异常？","正确配置应包含：1) 设置有效的 pineconeApiKey 和 pineconeEnv；2) 使用 build().init() 初始化客户端；3) 创建索引前确认 dimension 为 1536（适配 text-embedding-ada-002）；4) 调用 upsert 前确保索引已就绪。示例：PineconeClient.builder().pineconeApiKey(\"xxx\").pineconeEnv(\"us-west1-gcp-free\").requestTimeout(30).build().init()。",{"id":140,"question_zh":141,"answer_zh":142,"source_url":129},9168,"ChatGLM2 的 Java 实现是否支持多轮对话上下文裁剪以避免内存溢出？","当前 Java 版本未内置历史对话裁剪机制，但维护者建议使用 ConversationChain 内部管理历史，而非手动拼接。如需控制上下文长度，可在调用前对 history 列表进行截断（如保留最近 N 轮），避免传入过长历史导致模型超限。建议在业务层实现轮次限制逻辑。",{"id":144,"question_zh":145,"answer_zh":146,"source_url":119},9169,"如何为 langchain-java 项目贡献 JDK 8 兼容代码？","可基于 [jdk8 分支](https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Ftree\u002Fjdk8) 提交 PR。建议使用 GPT 工具将 JDK 17 特性（如文本块、var、Stream API）转换为 JDK 8 兼容语法，重点修改字符串处理和集合操作，保持原有逻辑不变。维护者欢迎社区协作维护 JDK 8 版本。",[148,153,158,163,168,173,178,183],{"id":149,"version":150,"summary_zh":151,"released_at":152},106589,"v0.2.2","## Features\r\n* [openai-client] Support create images #146\r\n\r\n## Improvements\r\n* [openai-client] Optimize class naming for ChatMessage\r\n\r\n## Bugfix\r\n* [openai-client] Incorrect type for logprobs field in CompletionResp #144\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fcompare\u002Fv0.2.1...v0.2.2","2023-12-21T02:54:45",{"id":154,"version":155,"summary_zh":156,"released_at":157},106590,"v0.2.1","## Features\r\n* Supports functions for openai-client #133\r\n\r\n## Improvements\r\n* optimize Readme\r\n* Ignore unknown fields by default for pinecone-client #122\r\n\r\n## Bugfix\r\n* Fix dependency convergence errors #129\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fcompare\u002Fv0.2.0...v0.2.1","2023-12-05T01:55:05",{"id":159,"version":160,"summary_zh":161,"released_at":162},106591,"v0.2.0","## Features\r\n* Suppoert stream response by @kael-aiur #60 #104 \r\n* Support GoogleSearchAPIWrapper @HamaWhiteGG  #110\r\n* Support Support online API documentation, automatically executed via GitHub Actions @HamaWhiteGG  #102\r\n\r\n## Improvements\r\n* Pass Chain's input parameters to the kwargs parameter of the BaseTool#run method. @HamaWhiteGG \r\n* Optimize the log output in the com.hw.langchain.agents.agent package. @HamaWhiteGG \r\n* Add more unit tests to SparkSqlToolkitTest. @HamaWhiteGG \r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fcompare\u002Fv0.1.12...v0.2.0","2023-10-07T16:38:29",{"id":164,"version":165,"summary_zh":166,"released_at":167},106592,"v0.1.12","## Features\r\n* Support Azure OpenAI endpoint by @wangtl #33\r\n* Support Ollama by #74\r\n* Support ChatGLM2 #74\r\n* Support Summarization(Stuff Chain)  #97\r\n* Support maxRetries for OpenAI and ChatOpenAI  #96\r\n* Support delete by ids for Pinecone #83\r\n\r\n## Improvements\r\n* Change the log level of the langchain-example module to DEBUG.\r\n\r\n## Bugfix\r\n* UncrecognizedPropertyException in CompetionRes #64\r\n\r\n\r\n## New Contributors\r\n* @wangtl made their first contribution in https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fpull\u002F92\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fcompare\u002Fv0.1.11...v0.1.12","2023-09-10T13:09:16",{"id":169,"version":170,"summary_zh":171,"released_at":172},106593,"v0.1.11","## Features\r\n* Support Flink SQL Agent by @HamaWhiteGG #56\r\n* Support Milvus #38 by @HamaWhiteGG #38 \r\n\r\n## Bugfix\r\n* Exception at BaseCombineDocumentsChain when use BaseRetrievalQA, as missing key [question] by @kael-aiur in https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fpull\u002F58\r\n* Hold \"warning\" filed when test in deprecated model like \"text-davinci-003\" by @wangmiao-1981 in https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fpull\u002F66\r\n\r\n## New Contributors\r\n* @kael-aiur made their first contribution in https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fpull\u002F58\r\n* @sandiegoe made their first contribution in https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fpull\u002F61\r\n* @wangmiao-1981 made their first contribution in https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fpull\u002F66\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fcompare\u002Fv0.1.10...v0.1.11","2023-08-04T02:27:06",{"id":174,"version":175,"summary_zh":176,"released_at":177},106594,"v0.1.10","## Features\r\n* Support i18n(pt-BR) for prompts in SQLDatabaseChain by @fwborges \r\n* Support Database memory by @zhangxiaojiawow \r\n* Support Spark SQL Agent by @HamaWhiteGG #53\r\n* Support API chains by @HamaWhiteGG #26\r\n* Support DirectoryLoadery @HamaWhiteGG #24\r\n* Support Load limited messages from chat memory by @zhangxiaojiawow \r\n\r\n## Improvements\r\n* Add okhttp interceptor by @zhangxiaojiawow\r\n* Add username and password for proxy by @HamaWhiteGG #30\r\n\r\n## New Contributors\r\n* @zhangxiaojiawow made their first contribution in https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fpull\u002F42\r\n* @fwborges made their first contribution in https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fpull\u002F45\r\n","2023-07-22T14:59:07",{"id":179,"version":180,"summary_zh":181,"released_at":182},106595,"v0.1.9","## What's Changed\r\n* Support MarkdownHeaderTextSplitter\r\n* Support Context aware text splitting and QA \u002F Chat\r\n\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fcompare\u002Fv0.1.8...v0.1.9","2023-07-13T12:20:53",{"id":184,"version":185,"summary_zh":186,"released_at":187},106596,"v0.1.8","## Features\r\n* Support OpenAIEmbeddings\r\n* Support Pinecone and add pinecone-client\r\n* Support Retrieval QA\r\n* Support Structured output parser (#3)\r\n* Support Chat Modes for Prompt Templates, Chains, Agents and Memory\r\n\r\n## Improvements\r\n* Add langchain-example module(Delete old QuickStart in langchain-core)\r\n* Update README.md\r\n\r\n## New Contributors\r\n* @tokuhirom made their first contribution in https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fpull\u002F20\r\n* @ThreaT made their first contribution in https:\u002F\u002Fgithub.com\u002FHamaWhiteGG\u002Flangchain-java\u002Fpull\u002F28","2023-07-06T12:37:45"]