diff --git a/.github/workflows/compile-and-deploy.yml b/.github/workflows/compile-and-deploy.yml deleted file mode 100644 index 041a03e..0000000 --- a/.github/workflows/compile-and-deploy.yml +++ /dev/null @@ -1,42 +0,0 @@ -name: 编译并部署 - -on: - workflow_dispatch: - -jobs: - compile-and-deploy: - runs-on: ubuntu-latest - steps: - - name: 检出 - uses: actions/checkout@v4 - - name: 编译 - run: | - sudo apt-get update - sudo apt-get install -y --no-install-recommends zip - docker run --rm -v ${PWD}:/docs registry-hk.fit2cloud.com/public/mkdocs-material:latest build - zip -r -q -9 site.zip site - ls -la site.zip - - name: 上传 - uses: appleboy/scp-action@v0.1.7 - with: - host: ${{ secrets.MAXKB_DOCS_HOST }} - username: ${{ secrets.MAXKB_DOCS_HOST_USERNAME }} - port: 22 - key: ${{ secrets.MAXKB_DOCS_HOST_KEY }} - source: "site.zip" - target: /tmp - - name: 部署 - uses: appleboy/ssh-action@v1.0.3 - with: - host: ${{ secrets.MAXKB_DOCS_HOST }} - username: ${{ secrets.MAXKB_DOCS_HOST_USERNAME }} - key: ${{ secrets.MAXKB_DOCS_HOST_KEY }} - port: 22 - script: | - cd /tmp - rm -rf site - unzip site.zip - rm -rf /opt/maxkb/docs/* - mv /tmp/site/* /opt/maxkb/docs - /opt/nginx/bin/reload-nginx.sh - rm -rf /tmp/site* diff --git a/docs/changelog.md b/docs/changelog.md index 3290c8b..ed4db4c 100644 --- a/docs/changelog.md +++ b/docs/changelog.md @@ -1,645 +1,622 @@ -## 1 版本说明 +## 1 Version Information !!! Abstract "" - **版本号说明：** 像其它优秀开源项目一样，MaxKB 将保持每月发布一个新的功能版本，功能版本中如遇较为紧急或严重的 Bug，将及时推出 Bug 修复的小版本。 - **MaxKB 版本号命名规则为：** v 大版本.功能版本. Bug 修复版本，示例如下： + **Version Notes:** Like other excellent open-source projects, MaxKB will maintain a monthly release of new feature versions. If urgent or serious bugs are encountered in a feature version, a bug fix minor version will be released promptly. + **MaxKB version naming rule:** v Major.Feature.Bug Fix, examples as follows: - - v1.0.1 是 v1.0.0 之后的 Bug 修复版本 - - v1.1.0 是 v1.0.0 之后的功能版本 + - v1.0.1 is a bug fix version after v1.0.0 + - v1.1.0 is a feature version after v1.0.0 -## 2 更新内容 +## 2 Update Content ### v1.9.1 -2025 年 1 月 9 日 +January 9, 2025 -!!! Abstract "功能优化 :sunflower:" - - 知识库：文档向量化时支持仅向量化非成功的分段。 - - 应用：通过第三方应用进行对话的日志，优化为按天记录每个用户的对话日志。(X-Pack) - -!!! Abstract "问题修复 :palm_tree:" - - 应用：修复进行语音输入时，提示中图片加载失败的问题。 - - 应用：修复嵌入应用节点，设置变量参数后，保存不生效的问题。 - - 应用：修复部分网络环境下使用图片理解模型对话结束后，提示Connection aborted的问题。 - - 应用：修复判断器节点的执行条件设置为any时不生效的问题。 - - 应用：修复添加图片理解和图片生成模型时，不能自定义输入基础模型的问题。 - - 应用：修复部分情况下在执行详情中会遗漏执行节点的问题。 +!!! Abstract "Feature Optimization :sunflower:" + - Knowledge Base: Support vectorizing only non-successful segments during document vectorization. + - Application: Conversation logs through third-party applications are optimized to record each user's conversation logs daily. (X-Pack) +!!! Abstract "Bug Fixes :palm_tree:" + - Application: Fixed the issue of image loading failure in prompts during voice input. + - Application: Fixed the issue where setting variable parameters in embedded application nodes would not take effect after saving. + - Application: Fixed the "Connection aborted" prompt issue after ending a conversation using the image understanding model in certain network environments. + - Application: Fixed the issue where the execution condition of the judgment node doesn't take effect when set to "any". + - Application: Fixed the issue where custom input of base models is not possible when adding image understanding and generation models. + - Application: Fixed the issue where execution nodes would be missing in execution details under certain circumstances. ### v1.9.0 -2025 年 1 月 2 日 - -!!! Abstract "新增功能 :star2:" - - 应用：新增图片生成节点。 - - 应用：支持上传音频文件，新增语音转文本节点。 - - 应用：新增文本转语音节点。 - - 应用：支持导出和导入应用。 - - 应用：工作流节点支持设置执行条件。 - - 应用：支持接入企业微信的微信客服。（X-Pack） - - 应用：公众号、企业微信、微信客服支持语音提问和语音回答。（X-Pack） - - 知识库：知识库和文档支持导出Excel文件与离线图片的ZIP包。 - - 知识库：上传文档选择文本文件类型时，新增XLS、XLSX、CSV、ZIP文件格式。 - - 知识库：上传文档选择QA问答对类型时，新增ZIP文件格式。 - - 模型设置：创建模型时，支持设置模型参数。 - - 模型设置：图片理解和图片生成模型支持设置模型参数。 - - 模型设置：图片理解模型新增支持Xinference、Ollama、豆包、阿里云百炼、Azure OpenAI、Gemini供应商。 - - 模型设置：图片生成模型新增支持Xinference、OpenAI、腾讯混元、通义千问、智谱AI、豆包、阿里云百炼、Azure OpenAI供应商。 - - 模型设置：向量模型新增支持Azure OpenAI供应商。 - - 模型设置：语音识别模型新增支持Azure OpenAI供应商。 - - 模型设置：语音合成模型新增支持Azure OpenAI供应商。 - -!!! Abstract "功能优化 :sunflower:" - - 应用：基本信息节点中的“用户输入“参数，添加参数时支持设置“显示默认值”选项。 - - 应用：表单收集节点的参数支持设置“显示默认值”选项。 - - 应用：对话URL支持携带“question=问题”参数，打开对话页面时自动发送问题。 - - 应用：上传图片时自动生成“请解析图片内容”的问题。 - - 应用：优化工作流节点的执行效率。 - - 知识库：文档列表支持批量取消向量化和批量取消生成问题。 - - 模型设置：Amazon Bedrock供应商添加大语言模型时，支持ProxyURL参数。 - -!!! Abstract "问题修复 :palm_tree:" - - 安全：修复函数库模块中远程命令执行的安全漏洞（CVE-2024-56137）。 - - 应用：修复浮窗对话框中不显示“新建对话”按钮的问题。 - - 应用：修复浮窗对话框右上角的icon颜色不一致的问题。 - - 应用：修复历史应用进行对话时会提示“缺少上下文类型”错误的问题。 - - 应用：修复使用低版本浏览器进行对话时报错的问题。 - - 应用：修复执行复杂工作流时，在部分情况下会漏执行节点的问题。 +January 2, 2025 + +!!! Abstract "New Features :star2:" + - Application: Added image generation node. + - Application: Support for uploading audio files, added speech-to-text node. + - Application: Added text-to-speech node. + - Application: Support for exporting and importing applications. + - Application: Workflow nodes support setting execution conditions. + - Application: Support for integrating with WeChat Work customer service. (X-Pack) + - Application: Official accounts, WeChat Work, and WeChat customer service support voice questions and answers. (X-Pack) + - Knowledge Base: Knowledge bases and documents support exporting ZIP packages with Excel files and offline images. + - Knowledge Base: When uploading documents and selecting text file type, added XLS, XLSX, CSV, and ZIP file formats. + - Knowledge Base: When uploading documents and selecting QA pairs type, added ZIP file format. + - Model Settings: When creating models, support for setting model parameters. + - Model Settings: Image understanding and image generation models support setting model parameters. + - Model Settings: Image understanding models added support for Xinference, Ollama, Doubao, Alibaba Cloud Bailian, Azure OpenAI, and Gemini providers. + - Model Settings: Image generation models added support for Xinference, OpenAI, Tencent Hunyuan, Tongyi Qianwen, Zhipu AI, Doubao, Alibaba Cloud Bailian, and Azure OpenAI providers. + - Model Settings: Vector models added support for Azure OpenAI provider. + - Model Settings: Speech recognition models added support for Azure OpenAI provider. + - Model Settings: Speech synthesis models added support for Azure OpenAI provider. + +!!! Abstract "Feature Optimization :sunflower:" + - Application: In the basic information node, the "user input" parameter supports setting the "show default value" option when adding parameters. + - Application: Parameters in the form collection node support setting the "show default value" option. + - Application: Conversation URL supports carrying the "question=query" parameter to automatically send the question when opening the conversation page. + - Application: Automatically generating "Please analyze the image content" question when uploading an image. + - Application: Optimized the execution efficiency of workflow nodes. + - Knowledge Base: Document list supports batch cancellation of vectorization and batch cancellation of question generation. + - Model Settings: When adding large language models for Amazon Bedrock provider, supports ProxyURL parameter. + +!!! Abstract "Bug Fixes :palm_tree:" + - Security: Fixed remote command execution security vulnerability in the function library module (CVE-2024-56137). + - Application: Fixed the issue where the "New Conversation" button is not displayed in the floating dialog box. + - Application: Fixed the issue of inconsistent icon colors in the upper right corner of the floating dialog box. + - Application: Fixed the issue where historical applications would prompt "missing context type" error during conversations. + - Application: Fixed the error when using older browser versions for conversations. + - Application: Fixed the issue where nodes would be missed when executing complex workflows under certain circumstances. ### v1.8.1 -2024 年 12 月 12 日 +December 12, 2024 -!!! Abstract "新增功能 :star2:" - - 应用：AI对话节点的历史聊天记录可设置工作流和节点的上下文。 - - 模型管理：支持智谱AI供应商的图片理解模型。 +!!! Abstract "New Features :star2:" + - Application: Historical chat records in AI conversation nodes can be set as context for workflows and nodes. + - Model Management: Support for Zhipu AI provider's image understanding model. -!!! Abstract "功能优化 :sunflower:" - - 应用：语音播放时过滤表单数据。 - - 系统：优化部分页面的显示样式。 +!!! Abstract "Feature Optimization :sunflower:" + - Application: Filter form data during voice playback. + - System: Optimized display styles of certain pages. -!!! Abstract "问题修复 :palm_tree:" - - 应用：按名称筛选结果后，不加载数据的问题。 - - 应用：修复浮窗图标尺寸过大时会被引导图遮盖的问题。 - - 知识库：修复文档按状态筛选结果错误的问题。 - - 知识库：修复拥有查看权限的成员可以导出文档的问题。 - - 知识库：修复拥有查看权限的成员可以修改分段启用/禁用状态的问题。 - - 知识库：修复上传无目录结构的PDF文件，只能识别第一页内容的问题。 - - 系统管理：修复Swagger文档中的Web站点知识库缺少参数的问题。（X-Pack） - - 系统管理：修复企业微信扫码登录配置错误的问题。（X-Pack） +!!! Abstract "Bug Fixes :palm_tree:" + - Application: Fixed the issue where data doesn't load after filtering results by name. + - Application: Fixed the issue where floating icon would be covered by the guide image when too large. + - Knowledge Base: Fixed the error in document filtering results by status. + - Knowledge Base: Fixed the issue where members with view permissions could export documents. + - Knowledge Base: Fixed the issue where members with view permissions could modify segment enable/disable status. + - Knowledge Base: Fixed the issue where only the first page content could be identified when uploading PDF files without directory structure. + - System Management: Fixed the missing parameter issue in the Web site knowledge base in Swagger documentation. (X-Pack) + - System Management: Fixed the WeChat Work scan code login configuration error. (X-Pack) ### v1.8.0 -2024 年 12 月 5 日 - -!!! Abstract "新增功能 :star2:" - - 应用：用户提问时支持上传文档和图片。 - - 应用：工作流新增文档内容提取节点。 - - 应用：工作流新增图片理解节点。 - - 应用：工作流新增表单收集节点。 - - 应用：工作流支持嵌入应用节点。 - - 应用：工作流画布支持收起或展开全部节点以及一键美化布局。 - - 应用：基础节点的用户输入参数中，支持多选框和选项卡组件。 - - 应用：开场白支持HTML渲染。 - - 应用：支持批量将对话日志添加到知识库。 - - 应用：支持按创建者显示应用。 - - 知识库：支持按创建者显示知识库。 - - 函数库：支持按创建者显示函数。 - - 模型设置：支持通义千问供应商的图片理解模型。 - - 模型设置：支持OpenAI供应商的图片理解模型。 - - 模型设置：支持腾讯混元供应商的图片理解模型。 - - 模型设置：支持阿里云百炼供应商的大语言模型。 - - 系统管理：支持OAuth2单点登录方式。（X-Pack） - -!!! Abstract "功能优化 :sunflower:" - - 知识库：优化文档状态，显示向量化和生成问题的进度详情。 - - 应用：基础节点的用户输入参数中，单选框组件支持标签、选项值的设置方式。 - - 应用：对话详情页面对AI回复内容进行标注时，保留上一次所选的知识库文档。 - - 应用：优化对话日志列表的显示顺序，按最近对话时间降序排序。 - - 应用：优化打开对话详情时，默认显示最新对话记录。 - -!!! Abstract "问题修复 :palm_tree:" - - 知识库：修复PDF文档中若包含不能识别的特殊字符时，上传文档会报错的问题。 - - 知识库：修复部分情况下为“成功”状态的文档生成问题时，无法下发任务的问题。 - - 知识库：修复由于网站编码问题，导致web站点知识库无法获取文本内容的问题。 - - 系统管理：修复拥有知识库管理权限的成员，无法查看知识库的问题。 - - 系统管理：修复拥有应用查看权限的成员，能够修改应用接入配置的问题。（X-Pack） +December 5, 2024 + +!!! Abstract "New Features :star2:" + - Application: Support for uploading documents and images when users ask questions. + - Application: Added document content extraction node to workflow. + - Application: Added image understanding node to workflow. + - Application: Added form collection node to workflow. + - Application: Support for embedded application nodes in workflow. + - Application: Workflow canvas supports collapsing or expanding all nodes and one-click layout beautification. + - Application: In basic node's user input parameters, support for checkbox and tab components. + - Application: Welcome message supports HTML rendering. + - Application: Support for batch adding conversation logs to knowledge base. + - Application: Support for displaying applications by creator. + - Knowledge Base: Support for displaying knowledge bases by creator. + - Function Library: Support for displaying functions by creator. + - Model Settings: Support for Tongyi Qianwen provider's image understanding model. + - Model Settings: Support for OpenAI provider's image understanding model. + - Model Settings: Support for Tencent Hunyuan provider's image understanding model. + - Model Settings: Support for Alibaba Cloud Bailian provider's large language model. + - System Management: Support for OAuth2 single sign-on method. (X-Pack) + +!!! Abstract "Feature Optimization :sunflower:" + - Knowledge Base: Optimized document status to display detailed progress of vectorization and question generation. + - Application: In basic node's user input parameters, radio button component supports setting labels and option values. + - Application: When annotating AI reply content on the conversation details page, retain the previously selected knowledge base document. + - Application: Optimized the display order of conversation logs list, sorted by latest conversation time in descending order. + - Application: Optimized opening conversation details to display the latest conversation record by default. + +!!! Abstract "Bug Fixes :palm_tree:" + - Knowledge Base: Fixed the error when uploading documents containing unrecognizable special characters in PDF documents. + - Knowledge Base: Fixed the issue where tasks couldn't be dispatched when generating questions for documents in "Success" status under certain circumstances. + - Knowledge Base: Fixed the issue where text content couldn't be obtained from website knowledge bases due to website encoding problems. + - System Management: Fixed the issue where members with knowledge base management permissions couldn't view knowledge bases. + - System Management: Fixed the issue where members with application view permissions could modify application access configuration. (X-Pack) ### v1.7.2 -2024 年 11 月 15 日 - -!!! Abstract "问题修复 :palm_tree:" - - 应用：修复显示设置中自定义上传的用户头像、AI回复头像图片显示变形的问题。 - - 应用：修复工作流中知识库检索节点收起后，无法发布应用的问题。 - - 应用：修复AI模型的参数设置中，若删除温度或最大输出 Tokens 参数，无法发布应用的问题。 - - 应用：修复应用中关联的知识库名称过长时，偶然不显示 tooltips 的问题。 - - 应用：修复应用列表中查询结果显示不正确的问题。 +November 15, 2024 +!!! Abstract "Bug Fixes :palm_tree:" + - Application: Fixed the issue where custom uploaded user avatars and AI reply avatars were distorted in display settings. + - Application: Fixed the issue where applications couldn't be published after collapsing the knowledge base retrieval node in the workflow. + - Application: Fixed the issue where applications couldn't be published if temperature or maximum output tokens parameters were deleted in AI model parameter settings. + - Application: Fixed the issue where tooltips occasionally wouldn't display when knowledge base names associated with applications were too long. + - Application: Fixed the issue where query results were displayed incorrectly in the application list. ### v1.7.1 -2024 年 11 月 8 日 - - -!!! Abstract "功能优化 :sunflower:" - - 应用：高级编排应用中保存节点的收起/展开状态。 - -!!! Abstract "问题修复 :palm_tree:" - - 应用：修复应用接入认证过的公众号，提问时不回复的问题。 - - 应用：修复应用接入公众号配置页面的样式问题。 - - 应用：函数中设置了非必填参数，在应用中引用该函数时，若未传值执行会报错的问题。 - - 应用：修复工作流中所有节点未打开返回内容时，无法查看执行详情的问题。 - - 应用：修复直接回答分段内容显示问题。 - - 应用：修复应用发布后，提问时未使用最新发布的应用配置的问题。 - - 知识库：修复导入DOCX文档时，在部分情况下无法识别图片问题。 - - 问答页面：修复AI回答时，若返回的MarkDown表格中含有图片内容会导致显示错位的问题。 - - 系统设置：修复团队成员的权限设置中，搜索后点击“全选”，会选中全部而不是搜索后的内容。 - - 模型管理：修复添加Xinference向量模型时，提示 API 域名无效的问题。 +November 8, 2024 +!!! Abstract "Feature Optimization :sunflower:" + - Application: Save the collapse/expand state of nodes in advanced orchestration applications. +!!! Abstract "Bug Fixes :palm_tree:" + - Application: Fixed the issue where authenticated official accounts wouldn't reply to questions. + - Application: Fixed style issues on the official account configuration page for application access. + - Application: Fixed the error that occurs during execution when non-required parameters are set in functions and those functions are referenced in applications without passing values. + - Application: Fixed the issue where execution details couldn't be viewed when no return content was enabled for all nodes in the workflow. + - Application: Fixed display issues with direct answer segment content. + - Application: Fixed the issue where the latest published application configuration wasn't used when asking questions after application publication. + - Knowledge Base: Fixed the issue where images couldn't be recognized when importing DOCX documents under certain circumstances. + - Q&A Page: Fixed the issue where image content in returned Markdown tables would cause display misalignment in AI answers. + - System Settings: Fixed the issue where clicking "Select All" after searching in team member permission settings would select all items instead of just the search results. + - Model Management: Fixed the issue where adding Xinference vector models would prompt "API domain name invalid". ### v1.7.0 -2024 年 10 月 31 日 - -!!! Abstract "新增功能 :star2:" - - 【知识库】支持知识库文档自动生成关联问题。 - - 【应用】工作流节点支持多进多出。 - - 【应用】支持设置语音播放的声音和语速。 - - 【应用】支持查看和恢复历史版本。 - - 【应用】支持设置对话日志的清除策略。 - - 【应用】高级编排应用支持自动保存设置，默认不开启，开启后默认5分钟保存一次。 - - 【应用】工作流编排界面在添加组件时支持按名称查询组件。 - - 【模型设置】支持自定义大语言模型和语音合成模型的参数。 - - 【模型设置】支持按模型类型、创建人、权限查询模型。 - - 【模型设置】模型面板中增加创建人信息。 - - 【模型设置】支持阿里云百炼的向量模型。 - - 【模型设置】支持 Amazon Bedrock 的向量模型。 - - 【模型设置】支持腾讯混元的向量模型。 - - 【模型设置】支持讯飞星火的向量模型。 - - 【模型设置】支持百度千帆的向量模型。 - - 【模型设置】支持阿里云百炼的语音识别和语音合成模型。 - - 【模型设置】支持 Xinference 的语音识别和语音合成模型。 - - 【问答页面】支持个性化显示设置。（X-Pack） - - 【问答页面】支持身份验证。（X-Pack） - - 【系统设置】支持企业微信、钉钉、飞书扫码登录。（X-Pack） - -!!! Abstract "功能优化 :sunflower:" - - 【应用】基础节点的全局变量优化为用户输入和接口传参两个参数表单。 - - 【应用】关联知识库时支持按名称查询知识库。 - - 【应用】关联知识库的名称过长时增加悬浮显示。 - - 【应用】知识库检索结果 和 满足直接回答的分段内容中包含分段标题。 - - 【应用】导出高级编排应用时包含知识库检索节点的引用分段数据。 - - 【应用】在语音输入时，控制台不再输出调试信息。 - - 【函数库】优化函数中 Python 编辑器组件，使得编写函数体验更友好。 - - 【团队管理】设置资源权限时增加全选按钮。 - - 【模型设置】创建模型时，支持按模型类型筛选供应商。 - - 【模型设置】优化供应商列表的显示顺序和分类，将供应商划分为公有模型供应商和私有模型供应商。 - - 【系统设置】优化修改密码的逻辑，先验证两次密码是否一致，再发邮件验证码。 - -!!! Abstract "问题修复 :palm_tree:" - - 【知识库】修复导入 License 文件后，会显示一次 License 过期的提示的问题。 - - 【知识库】修复当导出空知识库时，打开导出文件报错的问题。 - - 【知识库】修复当知识库文档中图片不存在时，AI 回复内容显示异常的问题。 - - 【知识库】修复在知识库上传 DOCX 文件后，部分情况下图片未正常导入的问题。 - - 【知识库】修复在知识库上传 PDF 文件，选择高级分段时，设置分段长度不生效的问题。 - - 【知识库】修复在知识库分段中插入 MarkDown 流程图，不显示文本内容的问题。 - - 【应用】修复当应用使用的模型为豆包智能体模型时，接入企业微信提问不回复的问题。 - - 【应用】修复知识来源引用分段的文件名过长时，界面显示错乱的问题。 - - 【应用】修复在接入的公众号中提问时，若回复内容在图片后面有文本内容, 无法返回图片的问题。 - - 【模型管理】修复添加 Xinference 模型时，如果模型名称和 ID 不一致，无法添加模型的问题。 - - 【模型管理】修复添加 Xinference 模型时，没有传认证参数导致提示 API 域名无效的问题。 - +October 31, 2024 + +!!! Abstract "New Features :star2:" + - [Knowledge Base] Support for auto-generating related questions for knowledge base documents. + - [Application] Workflow nodes support multiple inputs and outputs. + - [Application] Support for setting voice playback sound and speed. + - [Application] Support for viewing and restoring historical versions. + - [Application] Support for setting conversation log clearing policy. + - [Application] Advanced orchestration applications support auto-save settings, disabled by default, saves every 5 minutes when enabled. + - [Application] The workflow orchestration interface supports component name search when adding components. + - [Model Settings] Support for customizing parameters for large language models and speech synthesis models. + - [Model Settings] Support for querying models by model type, creator, and permissions. + - [Model Settings] Added creator information in the model panel. + - [Model Settings] Support for Alibaba Cloud Bailian vector models. + - [Model Settings] Support for Amazon Bedrock vector models. + - [Model Settings] Support for Tencent Hunyuan vector models. + - [Model Settings] Support for Xunfei Sparkdesk vector models. + - [Model Settings] Support for Baidu Qianfan vector models. + - [Model Settings] Support for Alibaba Cloud Bailian speech recognition and speech synthesis models. + - [Model Settings] Support for Xinference speech recognition and speech synthesis models. + - [Q&A Page] Support for personalized display settings. (X-Pack) + - [Q&A Page] Support for identity verification. (X-Pack) + - [System Settings] Support for WeChat Work, DingTalk, and Feishu scan code login. (X-Pack) + +!!! Abstract "Feature Optimization :sunflower:" + - [Application] Basic node's global variables optimized into two parameter forms: user input and interface parameters. + - [Application] Support for knowledge base name search when associating knowledge bases. + - [Application] Hover display for overly long knowledge base association names. + - [Application] Knowledge base retrieval results and direct answer segment content include segment titles. + - [Application] Exporting advanced orchestration applications includes referenced segment data from knowledge base retrieval nodes. + - [Application] No more debug information output in console during voice input. + - [Function Library] Optimized Python editor component in functions for a more friendly function writing experience. + - [Team Management] Added select all button when setting resource permissions. + - [Model Settings] Support for filtering providers by model type when creating models. + - [Model Settings] Optimized the display order and categorization of provider lists, dividing providers into public model providers and private model providers. + - [System Settings] Optimized password modification logic to verify password consistency first, then send email verification code. + +!!! Abstract "Bug Fixes :palm_tree:" + - [Knowledge Base] Fixed the issue where a license expiration notice would appear once after importing a License file. + - [Knowledge Base] Fixed the error when opening an exported file after exporting an empty knowledge base. + - [Knowledge Base] Fixed the abnormal display of AI reply content when images in knowledge base documents don't exist. + - [Knowledge Base] Fixed the issue where images weren't properly imported after uploading DOCX files to the knowledge base in certain cases. + - [Knowledge Base] Fixed the issue where segment length settings didn't take effect when uploading PDF files to the knowledge base with advanced segmentation. + - [Knowledge Base] Fixed the issue where inserting MarkDown flowcharts in knowledge base segments wouldn't display text content. + - [Application] Fixed the issue where questions in WeChat Work wouldn't receive replies when the application used Doubao intelligent agent models. + - [Application] Fixed the UI disorder issue when referenced segment file names in knowledge sources were too long. + - [Application] Fixed the issue where images couldn't be returned when text content appeared after images in replies to questions in the connected official account. + - [Model Management] Fixed the issue where models couldn't be added when model name and ID were inconsistent when adding Xinference models. + - [Model Management] Fixed the issue where no authentication parameters were passed causing "API domain name invalid" prompts when adding Xinference models. ### v1.6.1 -2024 年 09 月 29 日 +September 29, 2024 -!!! Abstract "问题修复 :palm_tree:" - - 【应用】修复使用AWS模型进行对话时报错的问题。 - - 【应用】修复升级后第一次进入高级编排应用编辑页面，显示空白的问题。 - - 【应用】修复高级编排应用设置全局变量后，打开调试对话框，选择变量后调试对话框会被关闭的问题。 - - 【应用】修复关联知识库中，知识库名称过长导致显示不全的问题。 - - 【知识库】修复分段内容超过单元格最大长度时，导出报错的问题。 +!!! Abstract "Bug Fixes :palm_tree:" + - [Application] Fixed the error when conversing using AWS models. + - [Application] Fixed the blank display issue when entering the advanced orchestration application edit page for the first time after upgrading. + - [Application] Fixed the issue where the debug dialog would close after selecting variables in the global variable settings of advanced orchestration applications. + - [Application] Fixed the incomplete display of knowledge base names that are too long in knowledge base associations. + - [Knowledge Base] Fixed the error when exporting when segment content exceeds the maximum length of a cell. ### v1.6.0 -2024 年 09 月 27 日 - -!!! Abstract "新增功能 :star2:" - - 【应用】支持语音方式问答。 - - 【应用】支持使用重排模型进行多路召回。 - - 【应用】支持自定义全局变量。 - - 【应用】支持标准 OpenAI 格式对话。 - - 【应用】支持渲染 ECharts 图表和 HTML 页面。 - - 【知识库】支持上传 Excel 和 CSV 表格文档。 - - 【知识库】支持对多个文档进行批量重新向量化操作。 - - 【知识库】支持问题批量关联分段。 - - 【函数库】】支持自定义设置函数权限与状态。 - - 【模型管理】】支持对接豆包、讯飞星火、OpenAI 供应商的语音识别、语音合成模型。 - - 【模型管理】】支持对接阿里云百炼、Xinference 以及本地模型的重排模型。 - - 【应用】支持一键接入企业微信、钉钉、飞书和微信公众号。（X-Pack） - - 【外观设置】支持自定义主题色、项目信息等设置。（X-Pack） - -!!! Abstract "功能优化 :sunflower:" - - 【知识库】PDF文档支持按目录章节进行分段。 - - 【应用】AI模型的参数设置中最大输出Tokens的最大值调整为10万，由用户自行控制大小。 - - 【应用】优化简易应用的设置，提示词分为无引用知识库提示词和引用知识库提示词。 - - 【应用】简易应用的问题优化调整至关联知识库的参数设置中，且支持设置提示词。 - - 【应用】简易应用的多轮对话调整为历史聊天记录，可以自定义设置历史记录数。 - - 【应用】简易应用调试预览的应用Logo可以自定义上传应用头像。 - - 【应用】高级编排应用的节点支持展开和收起，节省画布空间。 - - 【应用】高级编排应用点击节点右侧“+ ”号可直接添加节点。 - - 【应用】新增系统变量历史聊天记录和对话ID。 - - 【应用】对话日志详情中修改标注内容时支持Markdown样式编辑。 - - 【问答页面】优化换个答案按钮的图标。 - - 【用户管理】优化创建用户的用户名规则。 - - 【系统】优化弹窗只能点击“保存”或者“关闭”选项退出，避免编辑过程中误退出情况的发生。 - -!!! Abstract "问题修复 :palm_tree:" - - 【知识库】修复向MaxKB上传语雀导出的Word文档时，导入内容是空白的问题。 - - 【知识库】修复上传空白PDF文件报错的问题。 - - 【应用】修复开启多轮对话时，提问30秒以上不携带历史对话记录的问题。 - - 【应用】修复AI模型的参数设置中最大输出Tokens的提示描述错误的问题。 - - 【应用】修复通过应用的API Key进行对话时，若stream设置为false会报错的问题。 - - 【应用】修复在AI对话节点对调用的腾讯混元模型进行角色设定时，对话报错的问题。 - - 【团队成员】修复团队成员过多时会出现超出边界的问题。 +September 27, 2024 + +!!! Abstract "New Features :star2:" + - [Application] Support for voice Q&A. + - [Application] Support for multi-path retrieval using reranking models. + - [Application] Support for custom global variables. + - [Application] Support for standard OpenAI format conversations. + - [Application] Support for rendering ECharts charts and HTML pages. + - [Knowledge Base] Support for uploading Excel and CSV table documents. + - [Knowledge Base] Support for batch re-vectorization of multiple documents. + - [Knowledge Base] Support for batch association of questions with segments. + - [Function Library] Support for custom function permission and status settings. + - [Model Management] Support for speech recognition and speech synthesis models from Doubao, Xunfei Sparkdesk, and OpenAI providers. + - [Model Management] Support for reranking models from Alibaba Cloud Bailian, Xinference, and local models. + - [Application] One-click integration with WeChat Work, DingTalk, Feishu, and WeChat Official Accounts. (X-Pack) + - [Appearance Settings] Support for customizing theme colors, project information, and other settings. (X-Pack) + +!!! Abstract "Feature Optimization :sunflower:" + - [Knowledge Base] PDF documents support segmentation by directory chapters. + - [Application] The maximum value for tokens output in AI model parameter settings adjusted to 100,000, with size controlled by users. + - [Application] Optimized simple application settings, with prompts divided into prompts without referenced knowledge base and prompts with referenced knowledge base. + - [Application] Simple application questions optimized to knowledge base association parameter settings, with support for setting prompts. + - [Application] Simple application multi-turn conversation adjusted to chat history, with customizable history record count. + - [Application] Simple application debug preview's application logo can be customized with uploaded application avatars. + - [Application] Advanced orchestration application nodes support expansion and collapse to save canvas space. + - [Application] Clicking the "+" icon on the right side of nodes in advanced orchestration applications directly adds nodes. + - [Application] Added system variables for chat history and conversation ID. + - [Application] Support for Markdown style editing when modifying annotation content in conversation log details. + - [Q&A Page] Optimized the icon for the "try another answer" button. + - [User Management] Optimized username rules for creating users. + - [System] Optimized dialog boxes to only exit via "Save" or "Close" options, preventing accidental exits during editing. + +!!! Abstract "Bug Fixes :palm_tree:" + - [Knowledge Base] Fixed the issue where content is blank when uploading Word documents exported from Yuque to MaxKB. + - [Knowledge Base] Fixed the error when uploading blank PDF files. + - [Application] Fixed the issue where historical conversation records aren't carried when questions take more than 30 seconds with multi-turn conversation enabled. + - [Application] Fixed the error in description of maximum output tokens in AI model parameter settings. + - [Application] Fixed the error when conversing through an application's API Key with stream set to false. + - [Application] Fixed the error in conversations when setting role configurations for called Tencent Hunyuan models in AI conversation nodes. + - [Team Members] Fixed the issue where team members exceed boundaries when there are too many team members. ### v1.5.1 -2024 年 08 月 29 日 +August 29, 2024 -!!! Abstract "问题修复 :palm_tree:" - - 【应用】修复在应用设置中点击发布提示温度参数未设置的问题。 - - 【应用】修复使用本地模型进行提问时，获取tokens报错问题。 +!!! Abstract "Bug Fixes :palm_tree:" + - [Application] Fixed the issue of temperature parameter not set prompt when clicking publish in application settings. + - [Application] Fixed the error in obtaining tokens when asking questions using local models. ### v1.5.0 -2024 年 08 月 28 日 - -!!! Abstract "新增功能 :star2:" - - 【函数库】支持函数功能。 - - 【高级应用】工作流编排支持添加函数库中函数节点。 - - 【应用】支持复制应用。 - - 【应用】支持设置AI模型的temperature和max_tokens参数。 - - 【模型管理】支持对接豆包大模型。 - - 【模型管理】支持对接Amazon Bedrock 的大模型。 - - 【模型管理】支持对接腾讯混元大模型。 - - 【模型管理】支持对接 vLLM 大模型。 - - 【模型管理】支持对接 Xinference 大模型。 - - 【系统设置】支持 CAS 认证方式。（X-Pack） - - 【系统设置】支持 OIDC 认证方式。（X-Pack） - -!!! Abstract "功能优化 :sunflower:" - - 【知识库】优化文档向量化性能。 - - 【知识库】优化PDF文档的智能分段效果。 - - 【应用】优化应用名称过滤前后空格。 - -!!! Abstract "问题修复 :palm_tree:" - - 【系统】修复Django JSONField存在SQL注入漏洞（CVE-2024-42005）。 - - 【知识库】修复文档设置直接回答时相似度写0.000，提示设置成功，实际上不生效的问题。 - - 【知识库】修复智能分段时，单个段落超出10w字符报错的问题。 - - 【应用】修复对话日志导出报错的问题。 - - 【应用】修复编排页面AI对话节点提示词无法选中的问题。 - - 【应用】修复编排页面鼠标定位在AI对话节点的提示词中放大缩小事件冲突问题。 - - 【模型管理】修复模型名称前后端校验不统一问题。 - - 【模型管理】修复无法添加智谱AI模型的问题。 - - 【外观设置】修复登录图片设置上传大尺寸图片报错的问题。（X-Pack） - - 【外观设置】修复自定义外观设置后，还会显示默认外观的问题。（X-Pack） - +August 28, 2024 + +!!! Abstract "New Features :star2:" + - [Function Library] Support for function features. + - [Advanced Application] Workflow orchestration supports adding function nodes from the function library. + - [Application] Support for copying applications. + - [Application] Support for setting temperature and max_tokens parameters for AI models. + - [Model Management] Support for integrating with Doubao large models. + - [Model Management] Support for integrating with Amazon Bedrock large models. + - [Model Management] Support for integrating with Tencent Hunyuan large models. + - [Model Management] Support for integrating with vLLM large models. + - [Model Management] Support for integrating with Xinference large models. + - [System Settings] Support for CAS authentication method. (X-Pack) + - [System Settings] Support for OIDC authentication method. (X-Pack) + +!!! Abstract "Feature Optimization :sunflower:" + - [Knowledge Base] Optimized document vectorization performance. + - [Knowledge Base] Optimized intelligent segmentation effect for PDF documents. + - [Application] Optimized filtering out leading and trailing spaces in application names. + +!!! Abstract "Bug Fixes :palm_tree:" + - [System] Fixed Django JSONField SQL injection vulnerability (CVE-2024-42005). + - [Knowledge Base] Fixed the issue where setting direct answer similarity to 0.000 prompts successful setup but doesn't actually take effect. + - [Knowledge Base] Fixed the error when a single segment exceeds 100,000 characters during intelligent segmentation. + - [Application] Fixed the error when exporting conversation logs. + - [Application] Fixed the issue where prompts in AI conversation nodes cannot be selected on the orchestration page. + - [Application] Fixed the conflict between mouse zoom events and positioning in prompt fields in AI conversation nodes on the orchestration page. + - [Model Management] Fixed inconsistency in model name validation between frontend and backend. + - [Model Management] Fixed the issue where Zhipu AI models couldn't be added. + - [Appearance Settings] Fixed the error when uploading large-sized images for login image settings. (X-Pack) + - [Appearance Settings] Fixed the issue where default appearance is still displayed after customizing appearance settings. (X-Pack) + ### v1.4.1 -2024 年 08 月 08 日 +August 8, 2024 -!!! Abstract "新增功能 :star2:" - - 【应用】显示设置中自定义对话头像和浮窗入口图标支持GIF格式。（X-Pack） - - 【应用编排】指定回复和AI对话节点新增返回内容设置。 +!!! Abstract "New Features :star2:" + - [Application] Custom conversation avatars and floating entry icons in display settings support GIF format. (X-Pack) + - [Application Orchestration] Specified reply and AI conversation nodes added return content settings. -!!! Abstract "功能优化 :sunflower:" - - 【应用编排】优化开场白和提示词的编辑框支持放大。 - - 【问答页面】用户问题提问后显示换行格式。 +!!! Abstract "Feature Optimization :sunflower:" + - [Application Orchestration] Optimized welcome message and prompt editing boxes to support enlargement. + - [Q&A Page] User questions display line breaks after asking. -!!! Abstract "问题修复 :palm_tree:" - - 【应用】修复使用讯飞星火模型，当多用户同时提问后，回答内容会错乱的问题。 - - 【应用】修复在设置页面提问时，多轮对话设置不生效的问题。 - - 【应用】修复当应用只有使用权限，在高级编排页面一直提示没有权限的问题。 - - 【知识库】修复修改关联问题报错的问题。 - - 【知识库】修复添加分段时上传超大像素图谱报错的问题。 - - 【知识库】修复上传PDF问答后，部分内容丢失的问题。 - - 【知识库】创建Web站点知识库，选择向量模型不生效的问题。 - - 【知识库】修复上传QA问答对文件时，当表格中只有数字，获取数据失败的问题。 - - 【知识库】修复文档被删除后依然执行向量化的问题。 - - 【模型管理】修复模型名称设置超过20个字符报错的问题。 +!!! Abstract "Bug Fixes :palm_tree:" + - [Application] Fixed the issue where answer content would be jumbled when multiple users asked questions simultaneously using the Xunfei Sparkdesk model. + - [Application] Fixed the issue where multi-turn conversation settings don't take effect when asking questions on the settings page. + - [Application] Fixed the continuous permission prompt issue on the advanced orchestration page when an application only has usage permissions. + - [Knowledge Base] Fixed the error when modifying related questions. + - [Knowledge Base] Fixed the error when uploading ultra-high-pixel graphics when adding segments. + - [Knowledge Base] Fixed the issue of partial content loss after uploading PDF Q&A. + - [Knowledge Base] Fixed the issue where vector model selection doesn't take effect when creating Web site knowledge bases. + - [Knowledge Base] Fixed the issue of data retrieval failure when uploading QA pair files with only numbers in the table. + - [Knowledge Base] Fixed the issue of documents still being vectorized after deletion. + - [Model Management] Fixed the error when setting model names longer than 20 characters. ### v1.4.0 -2024 年 07 月 31 日 - -!!! Abstract "新增功能 :star2:" - - 【知识库】支持创建空知识库。 - - 【知识库】创建和设置知识库时支持选择不同的向量模型进行向量化。 - - 【模型管理】支持添加 OpenAI 、Ollama 和本地的向量模型。 - - 【模型管理】支持把模型权限设置为公用或私有。 - - 【应用】高级编排的指定回复节点支持快捷问题输出：`快捷问题`。 - - 【应用】支持设置浮窗模式的对话框入口图标。X-Pack） - - 【应用】支持自定义对话框的AI回复的头像。（X-Pack） - - 【应用】支持设置浮窗模式的对话框是否可拖拽位置。（X-Pack） - - 【应用】支持设置对话框是否显示历史记录。（X-Pack） - - 【系统设置】支持LDAP单点登录。（X-Pack） - - 【系统设置】支持自定义主题外观，设置系统的网站 logo、登录 logo、主题色、登录背景图、网站名称、欢迎语等。（X-Pack） - - 【系统设置】开放系统 API。（X-Pack） - -!!! Abstract "功能优化 :sunflower:" - - 【知识库】优化文档索引流程，提高知识点召回率。 - - 【知识库】调整分段内容最大长度为 10 万个字符。 - - 【应用】调整关联知识库时只能选择使用相同向量模型的知识库。 - - 【应用】显示知识来源调整至显示设置中。 - - 【模型管理】优化模型列表样式。 - - 【关于】优化关于页面显示信息。 - -!!! Abstract "问题修复 :palm_tree:" - - 【知识库】修复爬取Web网站文档时碰到无法访问页面时中断爬取任务的问题。 - - 【知识库】修复分段详情中关联问题过多时滚动条显示问题。 - - 【模型管理】修复在Ollama供应商有下载中模型时，切换到其他供应商的模型也变为下载中的问题。 - - 【问答页面】修复回答时若内容中包含代码不展开显示的问题。 - - 【问答页面】修复删除新建对话报错的问题。 - - - +July 31, 2024 + +!!! Abstract "New Features :star2:" + - [Knowledge Base] Support for creating empty knowledge bases. + - [Knowledge Base] Support for selecting different vector models for vectorization when creating and setting up knowledge bases. + - [Model Management] Support for adding OpenAI, Ollama, and local vector models. + - [Model Management] Support for setting model permissions as public or private. + - [Application] Specified reply nodes in advanced orchestration support quick question output: `Quick Question`. + - [Application] Support for setting the dialog box entry icon in floating window mode. (X-Pack) + - [Application] Support for customizing the AI reply avatar in the dialog box. (X-Pack) + - [Application] Support for setting whether the dialog box in floating window mode can be dragged. (X-Pack) + - [Application] Support for setting whether the dialog box displays history. (X-Pack) + - [System Settings] Support for LDAP single sign-on. (X-Pack) + - [System Settings] Support for customizing theme appearance, setting system website logo, login logo, theme color, login background image, website name, welcome message, etc. (X-Pack) + - [System Settings] Open system API. (X-Pack) + +!!! Abstract "Feature Optimization :sunflower:" + - [Knowledge Base] Optimized document indexing process to improve knowledge point recall rate. + - [Knowledge Base] Adjusted maximum segment content length to 100,000 characters. + - [Application] Adjusted to only allow selection of knowledge bases using the same vector model when associating knowledge bases. + - [Application] Display knowledge source moved to display settings. + - [Model Management] Optimized model list style. + - [About] Optimized displayed information on the About page. + +!!! Abstract "Bug Fixes :palm_tree:" + - [Knowledge Base] Fixed the issue where crawling tasks are interrupted when encountering inaccessible pages when crawling Web site documents. + - [Knowledge Base] Fixed scrollbar display issue when there are too many related questions in segment details. + - [Model Management] Fixed the issue where models from other providers also change to downloading status when Ollama provider has models downloading. + - [Q&A Page] Fixed the issue where code content in answers doesn't expand and display. + - [Q&A Page] Fixed the error when deleting a new conversation. ### v1.3.1 -2024 年 07 月 16 日 - -!!! Abstract "问题修复 :palm_tree:" - - 【应用概览】打开 API 访问凭据的地址显示 404。 - - 【应用编排】判断器中删除其中一个条件分支后，再添加条件分支，前端顺序号未更新。 - - 【应用编排】判断器中的选择变量的滚动条不能拖动。 - - 【应用编排】执行详情中知识库检索结果的排序不对，未按相似度降序排序。 - - 【应用编排】指定回复的内容在任务流程中传递会返回空。 - - 【问答页面】当应用关联的知识库中含有禁用状态的文档时，提问后点换个答案不会换一批命中分段。 - - 【知识库】在线知识库爬取文档名超过 128 个字符报错。 - -### v1.3.0 +July 16, 2024 + +!!! Abstract "Bug Fixes :palm_tree:" + - [Application Overview] Opening API access credentials address shows 404. + - [Application Orchestration] After deleting one condition branch in the judgment node and adding a condition branch again, the frontend sequence number is not updated. + - [Application Orchestration] The scrollbar for selecting variables in the judgment node cannot be dragged. + - [Application Orchestration] The order of knowledge base retrieval results in execution details is incorrect, not sorted by similarity in descending order. + - [Application Orchestration] The content of specified replies returns empty when passed in the task flow. + - [Q&A Page] When the application's associated knowledge base contains disabled documents, clicking "Try another answer" doesn't provide a new batch of hit segments after asking a question. + - [Knowledge Base] Online knowledge base crawling reports an error when document names exceed 128 characters. -2024 年 07 月 04 日 +### v1.3.0 -!!! Abstract "新增功能 :star2:" - - 【应用】支持创建简易配置和高级编排两种类型的应用，高级编排应用支持自定义工作流。 - - 【应用】高级编排的应用AI对话节点支持角色设定。 - - 【应用】高级编排的应用AI对话节点支持自定义设置携带的历史聊天记录数。 - - 【应用】高级编排的应用支持获取当前时间的系统变量。 - - 【对话日志】高级编排应用支持查看每个节点的执行详情。 +July 4, 2024 -!!! Abstract "问题修复 :palm_tree:" - - 【知识库】修复上传QA问答对时阻塞问题。 - - 【知识库】修复高级分段开启自动清洗后会把所有的换行符去掉的问题。 - - 【应用】修复提示词输入Json格式的内容时，提问会报错的问题。 +!!! Abstract "New Features :star2:" + - [Application] Support for creating two types of applications: simple configuration and advanced orchestration, with advanced orchestration applications supporting custom workflows. + - [Application] AI conversation nodes in advanced orchestration support role settings. + - [Application] AI conversation nodes in advanced orchestration support custom settings for the number of historical chat records to carry. + - [Application] Advanced orchestration applications support system variables for obtaining current time. + - [Conversation Logs] Advanced orchestration applications support viewing execution details of each node. +!!! Abstract "Bug Fixes :palm_tree:" + - [Knowledge Base] Fixed blocking issue when uploading QA pairs. + - [Knowledge Base] Fixed the issue where automatic cleaning in advanced segmentation would remove all line breaks. + - [Application] Fixed the error when asking questions with JSON format content in prompts. ### v1.2.1 -2024 年 06 月 14 日 - -!!! Abstract "新增功能 :star2:" - - 【问答页面】历史聊天记录支持逻辑删除。 +June 14, 2024 -!!! Abstract "功能优化 :sunflower:" - - 【问答页面】对话时文本输入框根据输入文本内容自适应高度。 - - 【应用】最大引用字符数调整为 10 万。 - - 【应用】对话时问题的最大字符数调整为 10 万。 - - 【应用】知识来源引用分段为 0 时也可打开详情。 +!!! Abstract "New Features :star2:" + - [Q&A Page] Historical chat records support logical deletion. -!!! Abstract "问题修复 :palm_tree:" - - 【应用】解决应用 API 文档接口调用报错的问题。 - - 【应用】解决嵌入第三方全屏模式描述错误的问题。 - - 【知识库】解决导入带有表格样式的 md 文件分段后表格格式丢失的问题。 +!!! Abstract "Feature Optimization :sunflower:" + - [Q&A Page] Text input box adapts height according to input text content during conversations. + - [Application] Maximum reference character count adjusted to 100,000. + - [Application] Maximum character count for questions during conversations adjusted to 100,000. + - [Application] Details can be opened even when knowledge source referenced segments is 0. +!!! Abstract "Bug Fixes :palm_tree:" + - [Application] Resolved the error issue when calling the application API documentation interface. + - [Application] Resolved description error in embedded third-party full-screen mode. + - [Knowledge Base] Resolved the issue of table format loss after segmenting imported md files with table styles. ### v1.2.0 -2024 年 05 月 30 日 - -!!! Abstract "新增功能 :star2:" - - 【知识库】支持上传 Excel/CSV 格式的 QA 问答对，自动关联问题。 - - 【知识库】支持上传 HTML 格式的文本文件。 - - 【知识库】支持知识库和文档重新向量化操作。 - - 【知识库】支持导出知识库和文档。 - - 【问答页面】支持新建对话和查看历史对话。 - - 【问答页面】支持导出对话记录。 - -!!! Abstract "功能优化 :sunflower:" - - 【应用】显示知识来源的引用分段，支持 markdown 样式显示。 - - 【应用】提问后模型返回异常信息时也要记录到对话日志中。 - - 【应用】在对话日志详情中修正答案时，问题可以修改。 - - 【应用】引用分段数最大值调整为 100 。 - - 【应用】开场白最大字符数调整为 4096 。 - -!!! Abstract "问题修复 :palm_tree:" - - 【应用】解决开场白中无法显示 markdown 样式的表格。 - - 【问答页面】解决提问时出现【用户问题】此字段不能为空字符串的问题。 - - 【知识库】修复删除最后一个分段时报错的问题。 - - 【知识库】上传 md格式文档，一个分段内容超过 4096 个字符时，上传报错。 - - 【关于】修复刷新浏览器后不显示版本号的问题。 - +May 30, 2024 + +!!! Abstract "New Features :star2:" + - [Knowledge Base] Support for uploading Excel/CSV format QA pairs, automatic question association. + - [Knowledge Base] Support for uploading HTML format text files. + - [Knowledge Base] Support for re-vectorizing knowledge bases and documents. + - [Knowledge Base] Support for exporting knowledge bases and documents. + - [Q&A Page] Support for creating new conversations and viewing historical conversations. + - [Q&A Page] Support for exporting conversation records. + +!!! Abstract "Feature Optimization :sunflower:" + - [Application] Referenced segments in knowledge sources support markdown style display. + - [Application] Model exception returns should still be recorded in conversation logs after questions. + - [Application] Questions can be modified when correcting answers in conversation log details. + - [Application] Maximum number of referenced segments adjusted to 100. + - [Application] Maximum welcome message character count adjusted to 4096. + +!!! Abstract "Bug Fixes :palm_tree:" + - [Application] Resolved the issue where markdown style tables can't be displayed in welcome messages. + - [Q&A Page] Resolved the "This field cannot be an empty string" error for [User Question] when asking questions. + - [Knowledge Base] Fixed the error when deleting the last segment. + - [Knowledge Base] Error when uploading md format documents when a segment's content exceeds 4096 characters. + - [About] Fixed the issue where version number isn't displayed after refreshing the browser. ### v1.1.3 -2024 年 05 月 15 日 +May 15, 2024 -!!! Abstract "新增功能 :star2:" - - 【模型管理】支持对接 Gemini 供应商的模型。 - - 【模型管理】支持对接 DeepSeek 供应商的模型。 - - 【模型管理】OpenAI 供应商的基础模型增加 gpt-4o 模型。 +!!! Abstract "New Features :star2:" + - [Model Management] Support for integrating with Gemini provider models. + - [Model Management] Support for integrating with DeepSeek provider models. + - [Model Management] Added gpt-4o model to OpenAI provider's base models. -!!! Abstract "功能优化 :sunflower:" - - 【知识库】文档问题列表的多选框优化增加可点击范围。 +!!! Abstract "Feature Optimization :sunflower:" + - [Knowledge Base] Optimized clickable range of checkboxes in document question lists. -!!! Abstract "问题修复 :palm_tree:" - - 【知识库】文档删除或迁移分段后，字符数不会重新计算。 - - 【知识库】批量导入选中将标题添加关联问题,标题为空时报错 。 - - 【模型管理】Ollama 模型中没有已下载的模型时，在 MaxKB 添加 Ollama 模型报错。 - - 【模型管理】添加模型时，上次的表单未清空。 +!!! Abstract "Bug Fixes :palm_tree:" + - [Knowledge Base] Document character count not recalculated after deleting or migrating segments. + - [Knowledge Base] Error when title is empty during batch import with add title as related question option. + - [Model Management] Error when adding Ollama models in MaxKB when there are no downloaded models in Ollama. + - [Model Management] Form not cleared after adding a model. ### v1.1.2 -2024 年 05 月 10 日 - -!!! Abstract "新增功能 :star2:" - - 【知识库】上传文档时支持将标题添加为关联问题（适用标题是问题的问答对）。 - - 【知识库】支持分段迁移和批量迁移。 - - 【知识库】支持分段批量删除。 - - 【知识库】文档直接回答时支持相似度设置。 - - 【应用】 新增 API Key 跨域的设置。 +May 10, 2024 -!!! Abstract "功能优化 :sunflower:" - - 【知识库】上传文档的分段预览优化为懒加载方式，避免数据量大时浏览器崩溃。 +!!! Abstract "New Features :star2:" + - [Knowledge Base] Support for adding titles as related questions when uploading documents (suitable for Q&A pairs where titles are questions). + - [Knowledge Base] Support for segment migration and batch migration. + - [Knowledge Base] Support for batch deletion of segments. + - [Knowledge Base] Support for similarity settings in document direct answers. + - [Application] Added cross-origin settings for API Keys. -!!! Abstract "问题修复 :palm_tree:" - - 【知识库】修复文档改名无法点击保存的问题。 - - 【知识库】修复导入文档时在特殊情况下导入失败。 - - 【用户管理】修复在创建新用户时，用户名含有 0 时校验不通过的问题。 - - 【UI】修复ui文件打包报错的问题。 +!!! Abstract "Feature Optimization :sunflower:" + - [Knowledge Base] Optimized uploaded document segment preview to lazy loading to prevent browser crashes with large data volumes. +!!! Abstract "Bug Fixes :palm_tree:" + - [Knowledge Base] Fixed inability to save when renaming documents. + - [Knowledge Base] Fixed import failure in special cases when importing documents. + - [User Management] Fixed validation failure when usernames contain 0 when creating new users. + - [UI] Fixed UI file packaging errors. ### v1.1.1 -2024 年 04 月 30 日 -!!! Abstract "功能优化 :sunflower:" - - 【知识库】上传文档的超时时间延长为 1 小时。 +April 30, 2024 -!!! Abstract "问题修复 :palm_tree:" - - 【应用】修复未选择AI 模型时，开启问题优化提问报错的问题。 - - 【应用】修复未选择AI 模型时，开启多轮对话报错的问题。 - - 【知识库】修复文档上传失败时，点击重试按钮无效的问题。 - - 【知识库】修复相同类型的知识库直接文档无法迁移的问题。 +!!! Abstract "Feature Optimization :sunflower:" + - [Knowledge Base] Extended document upload timeout to 1 hour. +!!! Abstract "Bug Fixes :palm_tree:" + - [Application] Fixed error when asking questions with question optimization enabled without selecting an AI model. + - [Application] Fixed error with multi-turn conversations enabled without selecting an AI model. + - [Knowledge Base] Fixed invalid retry button when document upload fails. + - [Knowledge Base] Fixed inability to migrate documents between knowledge bases of the same type. ### v1.1.0 -2024 年 04 月 29 日 - -!!! Abstract "新增功能 :star2:" - - 【知识库】新增全文检索和混合检索两种检索模式。 - - 【知识库】新增命中处理方式设置，支持设置文档下的分段命中后可直接返回分段内容，也可让模型优化后返回答案。 - - 【知识库】新增文档迁移知识库功能。 - - 【知识库】上传word文档、添加/编辑分段内容时支持上传本地图片。 - - 【应用】当用户提问题时，若无引用知识库内容，可指定回复内容。 - - 【应用】新增用户问答页面显示知识来源设置。 - - 【应用】支持自定义应用的 logo，用户端浏览器标签显示自定义 logo和应用名称。 - -!!! Abstract "功能优化 :sunflower:" - - 【应用】创建/编辑应用时支持不选择模型。 - - 【问答页面】 AI 回答时优化流式输出，加快数据展示。 - - 【搜索框】搜索框可以支持一键清除搜索条件。 - - 【模型管理】添加模型表单优化提示语。 - -!!! Abstract "问题修复 :palm_tree:" - - 【应用】修复对话日志详情不按时间排序的问题。 - - 【应用】修复问答页面提问达到最大限制的时候返回内容重复。 - - 【应用】修复离线环境提问后由于获取不到 tokens 数导致回答有报错内容的问题。 - - 【应用】修复修改应用信息保存时会导致问答页面进入 404 的问题。 - - 【模型管理】API Key 字符长度过长时，添加模型报错。 - - 【问答页面】问题过长时，AI 回答后会有报错信息。 +April 29, 2024 + +!!! Abstract "New Features :star2:" + - [Knowledge Base] Added two retrieval modes: full-text retrieval and hybrid retrieval. + - [Knowledge Base] Added hit handling method settings, supporting direct return of segment content or model-optimized answers when document segments are hit. + - [Knowledge Base] Added document migration functionality. + - [Knowledge Base] Support for uploading local images when uploading Word documents or adding/editing segment content. + - [Application] Ability to specify reply content when user questions don't reference knowledge base content. + - [Application] Added display knowledge source settings for user Q&A page. + - [Application] Support for custom application logos, displaying custom logos and application names in user browser tabs. + +!!! Abstract "Feature Optimization :sunflower:" + - [Application] Support for not selecting a model when creating/editing applications. + - [Q&A Page] Optimized streaming output during AI responses for faster data display. + - [Search Box] Search boxes support one-click clearing of search criteria. + - [Model Management] Optimized prompt text in add model form. + +!!! Abstract "Bug Fixes :palm_tree:" + - [Application] Fixed conversation log details not sorting by time. + - [Application] Fixed repeated content returns when reaching maximum question limit on Q&A page. + - [Application] Fixed issue with error content in answers due to inability to get token counts in offline environments. + - [Application] Fixed issue where editing application information would cause Q&A page to enter 404. + - [Model Management] Error when adding models when API Key character length is too long. + - [Q&A Page] Error information appears after AI answers when questions are too long. ### v1.0.4 -2024 年 04 月 19 日 - -!!! Abstract "新增功能 :star2:" - - 【模型管理】支持对接智谱 AI 开发放平台的通用大模型。 - - 【模型管理】支持对接讯飞星火大模型。 - +April 19, 2024 -!!! Abstract "功能优化 :sunflower:" - - 【登录】当前用户使用默认密码登录后需要先修改密码。 - - 【应用】对话日志详情修正答案保存分段时，知识库和文档支持搜索。 +!!! Abstract "New Features :star2:" + - [Model Management] Support for integrating with Zhipu AI development platform's general models. + - [Model Management] Support for integrating with Xunfei Sparkdesk models. +!!! Abstract "Feature Optimization :sunflower:" + - [Login] Current users logging in with default passwords must change their passwords first. + - [Application] Knowledge bases and documents support search when saving segments during answer correction in conversation log details. ### v1.0.3 -2024 年 04 月 18 日 +April 18, 2024 -!!! Abstract "新增功能 :star2:" - - 【知识库】上传文档支持选择文件夹。 - - 【模型管理】兼容对接OpenAI 格式的模型。 - - 【模型管理】支持对接通义千问的大模型。 - - 【模型管理】支持对接Kimi大模型。 +!!! Abstract "New Features :star2:" + - [Knowledge Base] Support for selecting folders when uploading documents. + - [Model Management] Compatible with models conforming to OpenAI format. + - [Model Management] Support for integrating with Tongyi Qianwen models. + - [Model Management] Support for integrating with Kimi models. -!!! Abstract "功能优化 :sunflower:" - - 【问答页面】优化AI 输出答案的效率。 - - 【命中测试】优化未命中分段时的提示内容。 +!!! Abstract "Feature Optimization :sunflower:" + - [Q&A Page] Optimized efficiency of AI answer output. + - [Hit Test] Optimized prompt content when no segments are hit. ### v1.0.2 -2024 年 04 月 17 日 +April 17, 2024 +!!! Abstract "Feature Optimization :sunflower:" + - [Knowledge Base] Improved appearance of dropdown component when questions are long when adding related questions in segments. -!!! Abstract "功能优化 :sunflower:" - - 【知识库】在分段中添加关联问题时，当问题比较长时下拉组件显示不美观。 - -!!! Abstract "问题修复 :palm_tree:" - - 【知识库】修复文本文件中存在空字符（\0）时，上传文档报错。 - - 【知识库】在问题中关联分段，当文档为空时搜索分段标题会。 - - 【应用】修复问答页面用户提问时，只输入空格提问后回答错位的问题。 - - 【应用】问答页面在回答中...时，点击停止回答不生效。 +!!! Abstract "Bug Fixes :palm_tree:" + - [Knowledge Base] Fixed error when uploading documents containing null characters (\0) in text files. + - [Knowledge Base] Error when searching segment titles when the document is empty when associating segments with questions. + - [Application] Fixed misalignment issue in answers when users only input spaces in questions on the Q&A page. + - [Application] Stop answer button doesn't work when AI is responding with "...". ### v1.0.1 -2024 年 04 月 15 日 - -!!! Abstract "功能优化 :sunflower:" - - 【知识库】文档列表、分段标题、分段内容、问题列表查询时大小写不敏感。 +April 15, 2024 -!!! Abstract "问题修复 :palm_tree:" - - 【知识库】修复整体同步知识库报错。 - - 【知识库】修复添加分段时关联问题报错。 - - 【知识库】文档列表为空时，在问题列表点击关联分段会404。 - - 【应用】修复在应用的命中测试点击分段内容报没有权限访问的问题。 +!!! Abstract "Feature Optimization :sunflower:" + - [Knowledge Base] Case-insensitive queries for document lists, segment titles, segment content, and question lists. +!!! Abstract "Bug Fixes :palm_tree:" + - [Knowledge Base] Fixed error in overall knowledge base synchronization. + - [Knowledge Base] Fixed error when associating questions when adding segments. + - [Knowledge Base] 404 error when clicking associate segments in the question list when document list is empty. + - [Application] Fixed "no permission to access" error when clicking segment content in application hit tests. ### v1.0.0 -2024 年 04 月 11 日 - - -!!! Abstract "新增功能 :star2:" - - 【知识库】支持上传 DOCX、PDF 格式的文档。 - - 【知识库】新增问题库，支持把用户的问题梳理到问题库，关联文档中的分段，从而能更精确的匹配到相关知识点。 - - 【应用】新增用量统计功能，针对每个应用支持统计提问用户数据、新增提问用户数、提问数、 消费Tokens数和用户满意度。 - - 【应用】对话日志支持导出。 - - 【模型管理】支持对接 OpenAI 模型。 - -!!! Abstract "功能优化 :sunflower:" - - 【应用】概览中重新生成公开访问链接会导致嵌入第三方脚本失效，增加确认提示。 - - 【应用】浮窗模式嵌入第三方的入口 logo 更新。 - - 【知识库】文档分段详情页面点击返回时，保留分页信息。 - - 【知识库】上传的文档限制调整为 100 MB。 - -!!! Abstract "问题修复 :palm_tree:" - - 【知识库】修复获取在线知识库的更新日志内容为乱码的问题。 - - 【知识库】修复创建知识库后，进入设置页面无法修改知识库的问题。 - - 【知识库】修复上传不支持格式的文档报错的问题。 - - 【应用】修复开启问题优化后，优化后的问题返回为空字符串时未处理的问题。 - - 【应用】修复从对话日志详情中添加标注内容报错的问题。 - - 【应用】修复浮窗模式的 放大按钮与关闭按钮未对齐。 - - 【用户管理】修复创建用户时登录密码使用默认密码MaxKB@123.. 提示规则不对。 +April 11, 2024 + +!!! Abstract "New Features :star2:" + - [Knowledge Base] Support for uploading DOCX and PDF format documents. + - [Knowledge Base] Added question database, support for organizing user questions into the question database and associating them with segments in documents for more precise matching of relevant knowledge points. + - [Application] Added usage statistics function, supporting statistics on question user data, new question users, number of questions, consumed tokens, and user satisfaction for each application. + - [Application] Support for exporting conversation logs. + - [Model Management] Support for integrating with OpenAI models. + +!!! Abstract "Feature Optimization :sunflower:" + - [Application] Added confirmation prompt when regenerating public access links in overview would invalidate embedded third-party scripts. + - [Application] Updated entry logo for floating mode embedded in third parties. + - [Knowledge Base] Retained pagination information when returning from document segment details page. + - [Knowledge Base] Uploaded document limit adjusted to 100 MB. + +!!! Abstract "Bug Fixes :palm_tree:" + - [Knowledge Base] Fixed garbled content issue when getting online knowledge base update logs. + - [Knowledge Base] Fixed inability to modify knowledge base after creation when entering settings page. + - [Knowledge Base] Fixed error when uploading documents in unsupported formats. + - [Application] Fixed unhandled issue when optimized questions return empty strings with question optimization enabled. + - [Application] Fixed error when adding annotation content from conversation log details. + - [Application] Fixed misalignment of maximize and close buttons in floating mode. + - [User Management] Fixed incorrect rule prompt when creating users with default password MaxKB@123... ### v0.9.1 -2024 年 03 月 27 日 - - -!!! Abstract "新增功能 :star2:" - - 【模型管理】Ollama 供应商添加/修改模型时支持自动下载。 - - 【关于】增加显示项目地址、用户手册、论坛求助入口。 +March 27, 2024 -!!! Abstract "功能优化 :sunflower:" - - 【模型管理】】Azure OpenAI 的基础模型支持自定义设置，自定义时需要填写 api-version。 - - 【模型管理】】千帆大模型的基础模型支持自定义输入官方标准的大模型（非自行部署）。 +!!! Abstract "New Features :star2:" + - [Model Management] Support for automatic download when adding/modifying Ollama provider models. + - [About] Added display of project address, user manual, and forum help entry. -!!! Abstract "问题修复 :palm_tree:" - - - 【应用】问答提问时同一时间限制只能发出一个快捷问题。 - - 【应用】修复问题优化开启后大模型未返回优化结果时的报错问题。 - - 【知识库】修复上传的 TXT 文档编码格式问题导致内容为空的问题。 - - 【知识库】修复文档列表勾选文档后会自动取消的问题。 - - 【用户管理】修复删除用户报错。 - - 【用户管理】编辑用户禁止修改用户名。 - - 【用户管理】修复禁用用户可以登录。 - - 【用户管理】修改用户姓名超过长度时错误提示不对。 - - 【用户管理】修复无法修改 admin 用户信息的问题。 - - 【用户管理】修复修改用户状态后用户位置变动的问题。 +!!! Abstract "Feature Optimization :sunflower:" + - [Model Management] Azure OpenAI base models support custom settings, requiring api-version when customizing. + - [Model Management] Qianfan large models support custom input of official standard large models (not self-deployed). +!!! Abstract "Bug Fixes :palm_tree:" + - [Application] Limit to only one quick question at a time during Q&A. + - [Application] Fixed error issue when large models don't return optimization results with question optimization enabled. + - [Knowledge Base] Fixed empty content issue due to encoding format problems in uploaded TXT documents. + - [Knowledge Base] Fixed automatic deselection of documents after checking them in document list. + - [User Management] Fixed error when deleting users. + - [User Management] Prevented username modification when editing users. + - [User Management] Fixed disabled users being able to log in. + - [User Management] Incorrect error prompt when modified username exceeds length. + - [User Management] Fixed inability to modify admin user information. + - [User Management] Fixed user position change after modifying user status. ### v0.9.0 -2024 年 03 月 20 日 - - -!!! Abstract "新增功能 :star2:" +March 20, 2024 - - **开箱即用**：支持直接上传文档、自动爬取在线文档，支持文本自动拆分、向量化、RAG（检索增强生成），智能问答交互体验好。 - - **无缝嵌入**：支持零编码快速嵌入到第三方业务系统，让已有系统快速拥有智能问答能力，提高用户满意度。 - - **多模型支持**：支持Ollama、百度千帆大模型和 Azure OpenAI 的大语言模型。 +!!! Abstract "New Features :star2:" + - **Ready to Use**: Support for directly uploading documents, automatically crawling online documents, automatic text splitting, vectorization, RAG (Retrieval Augmented Generation), and intelligent Q&A with good interactive experience. + - **Seamless Integration**: Support for zero-code quick embedding into third-party business systems, enabling existing systems to quickly have intelligent Q&A capabilities, improving user satisfaction. + - **Multi-model Support**: Support for large language models from Ollama, Baidu Qianfan, and Azure OpenAI. diff --git a/docs/contact.md b/docs/contact.md deleted file mode 100644 index c2eb6cf..0000000 --- a/docs/contact.md +++ /dev/null @@ -1,20 +0,0 @@ -## 1 项目地址 - -!!! Abstract "" - - - [MaxKB](https://github.com/1Panel-dev/MaxKB/issues) ![MaxKB stars](https://img.shields.io/github/stars/1Panel-dev/MaxKB) - -## 2 社区论坛 - -!!! Abstract "" - 大家可以在 [飞致云开源社区论坛 MaxKB 板块](https://bbs.fit2cloud.com/c/mk/11) 搜索、咨询、讨论 MaxKB 以及智能问答系统相关问题，并发表自己的见解与想法。 - -## 3 联系方式 - -!!! Abstract "" - - - 官网：https://maxkb.cn - - 邮箱：support@fit2cloud.com - - 电话：400-052-0755 - - 微信交流群： - ![企业微信](){ width="135px" } diff --git a/docs/dev_manual/APIKey_chat.md b/docs/dev_manual/APIKey_chat.md index 447687b..4ba11f4 100644 --- a/docs/dev_manual/APIKey_chat.md +++ b/docs/dev_manual/APIKey_chat.md @@ -1,51 +1,51 @@ !!! Abstract "" - MaxKB 社区版提供系统应用 API 以及兼容 OpenAI 标准格式进行对话，专业版提供系统级别 API，详细情况见：[系统API ](../user_manual/X-Pack/system_API.md)。 + The MaxKB community edition provides system application APIs and supports conversation in a format compatible with the OpenAI standard. The professional edition offers system-level APIs. For more details, refer to: [System API](../user_manual/X-Pack/system_API.md). -## 1 应用API +## 1 Application API -### 1.1 打开API文档 +### 1.1 Open API Documentation !!! Abstract "" - 在应用信息中复制并访问 swagger 地址。 + Copy and access the swagger address from the application information. -![API doc](../img/dev/app_swaagger.png) +![API doc](../img/dev/app_swagger.png) + +### 1.2 API Key Authentication -### 1.2 API Key认证 - ![APIKEY](../img/dev/app_apikey.png) -### 1.3 获取应用信息 +### 1.3 Retrieve Application Information !!! Abstract "" - 调用 profile 接口，获取应用详细信息（应用 id、name等）。 - -![ profile](../img/dev/app_profile.png) + Call the profile interface to get detailed application information (such as application ID, name, etc.). + +![profile](../img/dev/app_profile.png) -### 1.4 打开会话 +### 1.4 Open a Session !!! Abstract "" - 调用打开会话接口，输入上述步骤获取的应用 id，打开会话并获取会话 id。 + Call the open session interface, enter the application ID obtained in the steps above, open a session, and obtain a session ID. ![appid](../img/dev/chat_open.png) -### 1.5 进行对话 +### 1.5 Conduct a Conversation !!! Abstract "" - 调用对话接口，输入会话 id 进行对话。 + Call the conversation interface, and enter the session ID to conduct a conversation. ![chatid](../img/dev/app_chat.png) -## 2 标准OpenAI 格式 +## 2 Standard OpenAI Format !!! Abstract "" - MaxKB 应用兼容 OpenAI API 格式，在OpenAI API 原有调用方式的基础上替换为 MaxKB 应用提供的 Base URL 以及 API Key 即可。 + The MaxKB application is compatible with the OpenAI API format. You can replace the original OpenAI API Base URL and API Key with those provided by the MaxKB application. ![API doc](../img/dev/openai_baseurl.png) !!! Abstract "" ``` - # 将url 和 Authorization 替换为 MaxKB 应用实际真实的 Base URL 和 API Key。 + # Replace the URL and Authorization with the actual Base URL and API Key from the MaxKB application. curl https://maxkb.fit2cloud.com/api/application/xxxxxxxx-8c56-11ef-a99e-0242ac140003/chat/completions \ -H "Content-Type: application/json" \ @@ -54,11 +54,9 @@ "model": "gpt-3.5-turbo", "messages": [ { - "role": "你是杭州飞致云信息科技有限公司旗下产品 MaxKB 知识库问答系统的智能小助手，你的工作是帮助 MaxKB 用户解答使用中遇到的问题，用户找你回答问题时，你要把主题放在 MaxKB 知识库问答系统身上。", - "content": "MaxKB 是什么？" + "role": "You are the intelligent assistant of MaxKB, a product under Hangzhou Fit2Cloud Technology. Your job is to assist MaxKB users in solving problems they encounter during usage. When users come to you with questions, make sure to focus on the MaxKB Q&A system.", + "content": "What is MaxKB?" } ] }' ``` - - diff --git a/docs/faq/GPU_runOllama.md b/docs/faq/GPU_runOllama.md index a40cb1f..504b178 100644 --- a/docs/faq/GPU_runOllama.md +++ b/docs/faq/GPU_runOllama.md @@ -1,48 +1,47 @@ -# 如何让Ollama使用GPU运行LLM模型 +# How to Run LLM Models on Ollama Using GPU !!! Abstract "" - 以 NVIDIA 为例，说明 Ollama GPU 模式运行大模型的具体操作。 + Using NVIDIA as an example, this guide explains the specific steps to run large models on Ollama in GPU mode. - -## 1 安装英伟达容器安装包 +## 1 Install NVIDIA Container Toolkit !!! Abstract "" - 以 Ubuntu 22.04 为例（其它系统请参考：[英伟达官方文档](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/arch-overview.html)） + Using Ubuntu 22.04 as an example (for other systems, please refer to the [NVIDIA official documentation](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/arch-overview.html)) - - 配置 apt 源 + - Configure the apt source ``` curl -fsSL https://nvidia.github.io/libnvidia-container/gpgkey | sudo gpg --dearmor -o /usr/share/keyrings/nvidia-container-toolkit-keyring.gpg \ && curl -s -L https://nvidia.github.io/libnvidia-container/stable/deb/nvidia-container-toolkit.list | \ sed 's#deb https://#deb [signed-by=/usr/share/keyrings/nvidia-container-toolkit-keyring.gpg] https://#g' | \ sudo tee /etc/apt/sources.list.d/nvidia-container-toolkit.list ``` - - 更新源 + - Update the source ``` sudo apt-get update ``` - - 安装工具包 + - Install the toolkit ``` sudo apt-get install -y nvidia-container-toolkit ``` -## 2 使用 GPU 运行 Ollama +## 2 Run Ollama Using GPU !!! Abstract "" ``` - # 以后台模式运行 ollama 容器，并允许容器访问主机上所有可用的 NVIDIA GPU + # Run the Ollama container in the background mode and allow the container to access all available NVIDIA GPUs on the host docker run --gpus all -d -v /opt/ai/ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama ``` -## 3 使用 Ollama 下载模型 +## 3 Download Models Using Ollama !!! Abstract "" ``` - # 在线下载并运行模型 + # Download and run models online docker exec -it ollama ollama run qwen:7b ``` -## 4 在 MaxKB 中添加 Ollama 模型 +## 4 Add Ollama Models in MaxKB !!! Abstract "" - 下载完成并运行模型服务后，即可在 MaxKB 中添加对应的模型并使用。 -![添加模型](../img/FAQ/addmodel.png) \ No newline at end of file + Once the models are downloaded and the model service is running, you can add the corresponding models in MaxKB and use them. +![Add Model](../img/FAQ/addmodel.png) diff --git a/docs/faq/MaxKB_VS_Dify.md b/docs/faq/MaxKB_VS_Dify.md new file mode 100644 index 0000000..50b7d60 --- /dev/null +++ b/docs/faq/MaxKB_VS_Dify.md @@ -0,0 +1,81 @@ +!!! Abstract "" + + MaxKB and Dify are both open-source projects based on large language model technology, with differences in product positioning and capabilities: + + - Different product positioning: Dify is positioned as a development platform for large model applications, falling into the middleware category; MaxKB is positioned as an intelligent Q&A assistant based on large models and RAG, serving as an out-of-the-box final application. + - Product capability comparison: The following table is a capability comparison with LangChain, Flowise and other products provided by Dify's official documentation; MaxKB is an application built on LangChain that fills the gaps in enterprise-level features such as Workflow and SSO that LangChain lacks. + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
FeaturesDify.AILangChainFlowiseOpenAI Assistant APIMaxKB(Built on LangChain)
Programming MethodAPI + Application-orientedPython CodeApplication-orientedAPI-orientedAPI + Application-oriented
Supported LLMsRich varietyRich varietyRich varietyOpenAI onlyRich variety
RAG Engine✅✅✅✅✅
Agent✅✅❌✅✅
Workflow✅❌✅❌✅
Observability✅✅❌❌✅
Enterprise Features (SSO/Access Control)✅❌❌❌✅
Self-hosted Deployment✅✅✅❌✅
diff --git a/docs/faq/Offline_install_OllamaModel.md b/docs/faq/Offline_install_OllamaModel.md index 40f4614..c2069d4 100644 --- a/docs/faq/Offline_install_OllamaModel.md +++ b/docs/faq/Offline_install_OllamaModel.md @@ -1,27 +1,26 @@ -# 如何使用Ollama离线部署LLM大语言模型 +# How to Deploy LLM Large Language Models Offline Using Ollama +!!! Abstract "" + Using the qwen:0.5b model as an example, this guide details the process and steps for deploying LLM large language models offline with Ollama. -!!! Abstract "" - 以 qwen:0.5b 模型为例，详细说明 Ollama 离线部署 LLM 大语言模型的过程和步骤。 - -### 1 下载模型 +### 1 Download the Model !!! Abstract "" - 访问 huggingface 并下载 qwen1_5-0_5b-chat-q5_k_m.gguf 模型文件。 + Visit Hugging Face and download the qwen1_5-0_5b-chat-q5_k_m.gguf model file. ``` https://huggingface.co/Qwen/Qwen1.5-0.5B-Chat-GGUF/tree/main ``` -![下载模型](../img/FAQ/downModel.png) +![Download Model](../img/FAQ/downModel.png) -### 2 上传模型 +### 2 Upload the Model !!! Abstract "" - 将下载好的 Qwen1.5-0.5B-Chat-GGUF 模型文件上传到 Ollama 所在服务器。 + Upload the downloaded Qwen1.5-0.5B-Chat-GGUF model file to the server where Ollama is located. -### 3 创建Ollama Modelfile +### 3 Create Ollama Modelfile !!! Abstract "" - 创建一个名为 Modelfile 的文件，内容如下： + Create a file named Modelfile with the following content: ``` FROM ./qwen1_5-0_5b-chat-q5_k_m.gguf @@ -34,25 +33,24 @@ PARAMETER stop "<|im_start|>" PARAMETER stop "<|im_end|>" ``` - 说明：不同模型的 Modelfile 内容不同，可参考 Ollama 官网 [参数设置](https://ollama.com/library/qwen:0.5b) 。 - -![模型参数模版](../img/FAQ/modelSetting.png) + Note: The content of the Modelfile varies for different models. You can refer to the Ollama official website for [parameter settings](https://ollama.com/library/qwen:0.5b). +![Model Parameter Template](../img/FAQ/modelSetting.png) -## 4 在 Ollama 中创建模型 +## 4 Create the Model in Ollama !!! Abstract "" - 执行以下命令，创建模型： + Execute the following command to create the model: ``` ollama create qwen:0.5b -f Modelfile ``` - 执行以下命令，确认模型创建成功： + Execute the following command to confirm successful model creation: ``` ollama list ``` -![ollama查看模型列表](../img/FAQ/ollamaList.png) +![View Model List in Ollama](../img/FAQ/ollamaList.png) -## 5 在 MaxKB 中添加已创建的私有模型 +## 5 Add the Created Private Model in MaxKB -![MaxKB中添加模型](../img/FAQ/MaxKBaddModel.png) \ No newline at end of file +![Add Model in MaxKB](../img/FAQ/MaxKBaddModel.png) diff --git a/docs/faq/maxkb_In_halo.md b/docs/faq/maxkb_In_halo.md index 7526b65..393316a 100644 --- a/docs/faq/maxkb_In_halo.md +++ b/docs/faq/maxkb_In_halo.md @@ -1,7 +1,7 @@ -# 将 MaxKB 小助手集成到 Halo 中 +# Integrate MaxKB Assistant into Halo !!! Abstract "" - [Halo](https://github.com/halo-dev/halo) [ˈheɪloʊ]，强大易用的开源建站工具。通过 MaxKB 小助手插件，可以将文章内容同步至 MaxKB 知识库，并可通过浮窗模式为网站提供智能问答。 - 有关 MaxKB 小助手插件的使用说明，请查看 [官方文档](https://www.halo.run/store/apps/app-aWHcE)。 + [Halo](https://github.com/halo-dev/halo) [ˈheɪloʊ] is a powerful and user-friendly open-source site-building tool. Through the MaxKB Assistant plugin, you can synchronize article content to the MaxKB knowledge base and provide intelligent Q&A on the website through a floating window mode. + For instructions on using the MaxKB Assistant plugin, please refer to the [official documentation](https://www.halo.run/store/apps/app-aWHcE). -![halo](../img/FAQ/halo.png) \ No newline at end of file +![halo](../img/FAQ/halo.png) diff --git a/docs/img/FAQ/MaxKBaddModel.png b/docs/img/FAQ/MaxKBaddModel.png index d437e4a..e4d00be 100644 Binary files a/docs/img/FAQ/MaxKBaddModel.png and b/docs/img/FAQ/MaxKBaddModel.png differ diff --git a/docs/img/FAQ/addmodel.png b/docs/img/FAQ/addmodel.png index b1d550b..99a2d83 100644 Binary files a/docs/img/FAQ/addmodel.png and b/docs/img/FAQ/addmodel.png differ diff --git a/docs/img/app/Access restrictions.png b/docs/img/app/Access restrictions.png deleted file mode 100644 index 30b8bbf..0000000 Binary files a/docs/img/app/Access restrictions.png and /dev/null differ diff --git a/docs/img/app/QA.png b/docs/img/app/QA.png index 2737612..e2620e6 100644 Binary files a/docs/img/app/QA.png and b/docs/img/app/QA.png differ diff --git a/docs/img/app/access_restrictions.png b/docs/img/app/access_restrictions.png new file mode 100644 index 0000000..6b07633 Binary files /dev/null and b/docs/img/app/access_restrictions.png differ diff --git a/docs/img/app/ai_chat.png b/docs/img/app/ai_chat.png index 633c915..7b1215d 100644 Binary files a/docs/img/app/ai_chat.png and b/docs/img/app/ai_chat.png differ diff --git a/docs/img/app/app-closed.png b/docs/img/app/app-closed.png index 415c8bd..4a7dabf 100644 Binary files a/docs/img/app/app-closed.png and b/docs/img/app/app-closed.png differ diff --git a/docs/img/app/app_auth.png b/docs/img/app/app_auth.png index 0d37471..2f4a379 100644 Binary files a/docs/img/app/app_auth.png and b/docs/img/app/app_auth.png differ diff --git a/docs/img/app/app_export.png b/docs/img/app/app_export.png index b5dc236..94a176f 100644 Binary files a/docs/img/app/app_export.png and b/docs/img/app/app_export.png differ diff --git a/docs/img/app/app_function.png b/docs/img/app/app_function.png index d52e51b..25416d3 100644 Binary files a/docs/img/app/app_function.png and b/docs/img/app/app_function.png differ diff --git a/docs/img/app/app_hittesting.png b/docs/img/app/app_hittesting.png index 22b017b..972cbda 100644 Binary files a/docs/img/app/app_hittesting.png and b/docs/img/app/app_hittesting.png differ diff --git a/docs/img/app/app_import.png b/docs/img/app/app_import.png index d82133b..a0976b8 100644 Binary files a/docs/img/app/app_import.png and b/docs/img/app/app_import.png differ diff --git a/docs/img/app/app_password1.png b/docs/img/app/app_password1.png index 699d27c..392def6 100644 Binary files a/docs/img/app/app_password1.png and b/docs/img/app/app_password1.png differ diff --git a/docs/img/app/app_password2.png b/docs/img/app/app_password2.png index 596e592..ed76514 100644 Binary files a/docs/img/app/app_password2.png and b/docs/img/app/app_password2.png differ diff --git a/docs/img/app/app_setting.png b/docs/img/app/app_setting.png index 406f47c..90d19ad 100644 Binary files a/docs/img/app/app_setting.png and b/docs/img/app/app_setting.png differ diff --git a/docs/img/app/app_start.png b/docs/img/app/app_start.png index acab05c..42d00b7 100644 Binary files a/docs/img/app/app_start.png and b/docs/img/app/app_start.png differ diff --git a/docs/img/app/app_subapp.png b/docs/img/app/app_subapp.png index 6e7c51b..9d4bc89 100644 Binary files a/docs/img/app/app_subapp.png and b/docs/img/app/app_subapp.png differ diff --git a/docs/img/app/app_workflow.png b/docs/img/app/app_workflow.png index 41f8481..0b91678 100644 Binary files a/docs/img/app/app_workflow.png and b/docs/img/app/app_workflow.png differ diff --git a/docs/img/app/audio2text.png b/docs/img/app/audio2text.png index e0a8ddb..a37cc5a 100644 Binary files a/docs/img/app/audio2text.png and b/docs/img/app/audio2text.png differ diff --git a/docs/img/app/auto_save.png b/docs/img/app/auto_save.png index 5011e07..b01d561 100644 Binary files a/docs/img/app/auto_save.png and b/docs/img/app/auto_save.png differ diff --git a/docs/img/app/condition_statement.png b/docs/img/app/condition_statement.png index 9cb8162..38eeedc 100644 Binary files a/docs/img/app/condition_statement.png and b/docs/img/app/condition_statement.png differ diff --git a/docs/img/app/dataset_search.png b/docs/img/app/dataset_search.png index 5e8ad36..ded128f 100644 Binary files a/docs/img/app/dataset_search.png and b/docs/img/app/dataset_search.png differ diff --git a/docs/img/app/defult_workflow.png b/docs/img/app/defult_workflow.png index e062af6..eb71f8d 100644 Binary files a/docs/img/app/defult_workflow.png and b/docs/img/app/defult_workflow.png differ diff --git a/docs/img/app/direct_reply.png b/docs/img/app/direct_reply.png index 415377c..3c0135a 100644 Binary files a/docs/img/app/direct_reply.png and b/docs/img/app/direct_reply.png differ diff --git a/docs/img/app/doc_extract.png b/docs/img/app/doc_extract.png index bdee416..13883c8 100644 Binary files a/docs/img/app/doc_extract.png and b/docs/img/app/doc_extract.png differ diff --git a/docs/img/app/embed.png b/docs/img/app/embed.png index 8482cd6..e831d84 100644 Binary files a/docs/img/app/embed.png and b/docs/img/app/embed.png differ diff --git a/docs/img/app/exec_detail.png b/docs/img/app/exec_detail.png index ccfa797..bb09389 100644 Binary files a/docs/img/app/exec_detail.png and b/docs/img/app/exec_detail.png differ diff --git a/docs/img/app/form.png b/docs/img/app/form.png index 4eb9ef6..989d108 100644 Binary files a/docs/img/app/form.png and b/docs/img/app/form.png differ diff --git a/docs/img/app/log_detail.png b/docs/img/app/log_detail.png index 78868e6..fa492ef 100644 Binary files a/docs/img/app/log_detail.png and b/docs/img/app/log_detail.png differ diff --git a/docs/img/app/log_details_edit.png b/docs/img/app/log_details_edit.png index 012f3c1..4ed3dc2 100644 Binary files a/docs/img/app/log_details_edit.png and b/docs/img/app/log_details_edit.png differ diff --git a/docs/img/app/loglist.png b/docs/img/app/loglist.png index 3684f9c..193f449 100644 Binary files a/docs/img/app/loglist.png and b/docs/img/app/loglist.png differ diff --git a/docs/img/app/node_condition.png b/docs/img/app/node_condition.png index f3c3e67..9c3641e 100644 Binary files a/docs/img/app/node_condition.png and b/docs/img/app/node_condition.png differ diff --git a/docs/img/app/picture_extract.png b/docs/img/app/picture_extract.png index 6ac4007..a0d2e41 100644 Binary files a/docs/img/app/picture_extract.png and b/docs/img/app/picture_extract.png differ diff --git a/docs/img/app/question_opt.png b/docs/img/app/question_opt.png index 34613ae..2282e08 100644 Binary files a/docs/img/app/question_opt.png and b/docs/img/app/question_opt.png differ diff --git a/docs/img/app/rerank.png b/docs/img/app/rerank.png index 4c06ca0..e48e5d7 100644 Binary files a/docs/img/app/rerank.png and b/docs/img/app/rerank.png differ diff --git a/docs/img/app/selectAppType.png b/docs/img/app/selectAppType.png index 92b163b..8fba6a5 100644 Binary files a/docs/img/app/selectAppType.png and b/docs/img/app/selectAppType.png differ diff --git a/docs/img/app/start_node.png b/docs/img/app/start_node.png index 5a6850b..f390920 100644 Binary files a/docs/img/app/start_node.png and b/docs/img/app/start_node.png differ diff --git a/docs/img/app/text2audio.png b/docs/img/app/text2audio.png index 73cd8e6..d5dbc52 100644 Binary files a/docs/img/app/text2audio.png and b/docs/img/app/text2audio.png differ diff --git a/docs/img/app/upload_config.png b/docs/img/app/upload_config.png index dfdee15..9c69bed 100644 Binary files a/docs/img/app/upload_config.png and b/docs/img/app/upload_config.png differ diff --git a/docs/img/app/user_input.png b/docs/img/app/user_input.png index 3b28fc7..a4bbcb9 100644 Binary files a/docs/img/app/user_input.png and b/docs/img/app/user_input.png differ diff --git a/docs/img/app/view_setting.png b/docs/img/app/view_setting.png index 5f5d3c2..d1ce094 100644 Binary files a/docs/img/app/view_setting.png and b/docs/img/app/view_setting.png differ diff --git a/docs/img/app/vision_gen.png b/docs/img/app/vision_gen.png index f219b91..8c5a718 100644 Binary files a/docs/img/app/vision_gen.png and b/docs/img/app/vision_gen.png differ diff --git a/docs/img/app/workflow.png b/docs/img/app/workflow.png index f20e52c..dafdcf3 100644 Binary files a/docs/img/app/workflow.png and b/docs/img/app/workflow.png differ diff --git a/docs/img/app/workflow_view.png b/docs/img/app/workflow_view.png index 74a6fac..28e1ec5 100644 Binary files a/docs/img/app/workflow_view.png and b/docs/img/app/workflow_view.png differ diff --git a/docs/img/dataset/DataEase_doc.png b/docs/img/dataset/DataEase_doc.png deleted file mode 100644 index 54ac490..0000000 Binary files a/docs/img/dataset/DataEase_doc.png and /dev/null differ diff --git a/docs/img/dataset/add_segmentation.png b/docs/img/dataset/add_segmentation.png index a7b1286..d5741b0 100644 Binary files a/docs/img/dataset/add_segmentation.png and b/docs/img/dataset/add_segmentation.png differ diff --git a/docs/img/dataset/advanced_segmentation.png b/docs/img/dataset/advanced_segmentation.png index 06222ce..5c1f1a0 100644 Binary files a/docs/img/dataset/advanced_segmentation.png and b/docs/img/dataset/advanced_segmentation.png differ diff --git a/docs/img/dataset/auto_paragraph.png b/docs/img/dataset/auto_paragraph.png index 56fd713..23ac359 100644 Binary files a/docs/img/dataset/auto_paragraph.png and b/docs/img/dataset/auto_paragraph.png differ diff --git a/docs/img/dataset/automatic_paragraphing.png b/docs/img/dataset/automatic_paragraphing.png index 405e93b..23ac359 100644 Binary files a/docs/img/dataset/automatic_paragraphing.png and b/docs/img/dataset/automatic_paragraphing.png differ diff --git a/docs/img/dataset/create.jpg b/docs/img/dataset/create.jpg index 1eac902..9fbbfd6 100644 Binary files a/docs/img/dataset/create.jpg and b/docs/img/dataset/create.jpg differ diff --git a/docs/img/dataset/create_offline_dataset.png b/docs/img/dataset/create_offline_dataset.png index 1e4e113..b4f0ea6 100644 Binary files a/docs/img/dataset/create_offline_dataset.png and b/docs/img/dataset/create_offline_dataset.png differ diff --git a/docs/img/dataset/create_problem.png b/docs/img/dataset/create_problem.png index 7820a05..aa2eeb4 100644 Binary files a/docs/img/dataset/create_problem.png and b/docs/img/dataset/create_problem.png differ diff --git a/docs/img/dataset/dataset_actions.png b/docs/img/dataset/dataset_actions.png index e3a992a..9d8512d 100644 Binary files a/docs/img/dataset/dataset_actions.png and b/docs/img/dataset/dataset_actions.png differ diff --git a/docs/img/dataset/dataset_del.png b/docs/img/dataset/dataset_del.png index c67cdd0..d420132 100644 Binary files a/docs/img/dataset/dataset_del.png and b/docs/img/dataset/dataset_del.png differ diff --git a/docs/img/dataset/dataset_embedding.png b/docs/img/dataset/dataset_embedding.png index a895d58..fcd0fd7 100644 Binary files a/docs/img/dataset/dataset_embedding.png and b/docs/img/dataset/dataset_embedding.png differ diff --git a/docs/img/dataset/dataset_export.png b/docs/img/dataset/dataset_export.png index 56666fe..5d878c8 100644 Binary files a/docs/img/dataset/dataset_export.png and b/docs/img/dataset/dataset_export.png differ diff --git a/docs/img/dataset/dataset_file_export.png b/docs/img/dataset/dataset_file_export.png index 5f61b4d..bb8ea97 100644 Binary files a/docs/img/dataset/dataset_file_export.png and b/docs/img/dataset/dataset_file_export.png differ diff --git a/docs/img/dataset/dataset_setting.png b/docs/img/dataset/dataset_setting.png index fd8ad33..926887e 100644 Binary files a/docs/img/dataset/dataset_setting.png and b/docs/img/dataset/dataset_setting.png differ diff --git a/docs/img/dataset/del_segmentation.png b/docs/img/dataset/del_segmentation.png index 31bd1f7..391d3ab 100644 Binary files a/docs/img/dataset/del_segmentation.png and b/docs/img/dataset/del_segmentation.png differ diff --git a/docs/img/dataset/doc_delete.png b/docs/img/dataset/doc_delete.png index c1f9665..eb97432 100644 Binary files a/docs/img/dataset/doc_delete.png and b/docs/img/dataset/doc_delete.png differ diff --git a/docs/img/dataset/doc_enable.png b/docs/img/dataset/doc_enable.png index 315c20d..b9dcfa5 100644 Binary files a/docs/img/dataset/doc_enable.png and b/docs/img/dataset/doc_enable.png differ diff --git a/docs/img/dataset/doc_setting.png b/docs/img/dataset/doc_setting.png index f10b5b5..445d642 100644 Binary files a/docs/img/dataset/doc_setting.png and b/docs/img/dataset/doc_setting.png differ diff --git a/docs/img/dataset/edit_segmentation.png b/docs/img/dataset/edit_segmentation.png index 1490a18..99f9cd6 100644 Binary files a/docs/img/dataset/edit_segmentation.png and b/docs/img/dataset/edit_segmentation.png differ diff --git a/docs/img/dataset/gen_question.png b/docs/img/dataset/gen_question.png index 1857070..a2f75a9 100644 Binary files a/docs/img/dataset/gen_question.png and b/docs/img/dataset/gen_question.png differ diff --git a/docs/img/dataset/hittesting.png b/docs/img/dataset/hittesting.png index 50f61c7..821095b 100644 Binary files a/docs/img/dataset/hittesting.png and b/docs/img/dataset/hittesting.png differ diff --git a/docs/img/dataset/move_segmentation.png b/docs/img/dataset/move_segmentation.png index 1de73b5..6e5dade 100644 Binary files a/docs/img/dataset/move_segmentation.png and b/docs/img/dataset/move_segmentation.png differ diff --git a/docs/img/dataset/move_web_doc.png b/docs/img/dataset/move_web_doc.png index a45fa59..8d4faa4 100644 Binary files a/docs/img/dataset/move_web_doc.png and b/docs/img/dataset/move_web_doc.png differ diff --git a/docs/img/dataset/preview_segmentation.png b/docs/img/dataset/preview_segmentation.png index ac99679..cae4692 100644 Binary files a/docs/img/dataset/preview_segmentation.png and b/docs/img/dataset/preview_segmentation.png differ diff --git a/docs/img/dataset/problem_detail.png b/docs/img/dataset/problem_detail.png index 9cfbe93..3c88193 100644 Binary files a/docs/img/dataset/problem_detail.png and b/docs/img/dataset/problem_detail.png differ diff --git a/docs/img/dataset/problem_list.png b/docs/img/dataset/problem_list.png index 3eaf0f9..be6cea9 100644 Binary files a/docs/img/dataset/problem_list.png and b/docs/img/dataset/problem_list.png differ diff --git a/docs/img/dataset/problem_segmentation.png b/docs/img/dataset/problem_segmentation.png index 349f727..6893583 100644 Binary files a/docs/img/dataset/problem_segmentation.png and b/docs/img/dataset/problem_segmentation.png differ diff --git a/docs/img/dataset/processing.png b/docs/img/dataset/processing.png index b2afeb7..41aef8a 100644 Binary files a/docs/img/dataset/processing.png and b/docs/img/dataset/processing.png differ diff --git a/docs/img/dataset/segmentation_management.png b/docs/img/dataset/segmentation_management.png index f454f12..412a403 100644 Binary files a/docs/img/dataset/segmentation_management.png and b/docs/img/dataset/segmentation_management.png differ diff --git a/docs/img/dataset/sysn_dataset.png b/docs/img/dataset/sysn_dataset.png index 581feb8..3368a0e 100644 Binary files a/docs/img/dataset/sysn_dataset.png and b/docs/img/dataset/sysn_dataset.png differ diff --git a/docs/img/dataset/sysn_web_doc.png b/docs/img/dataset/sysn_web_doc.png index 57d71e9..3efa527 100644 Binary files a/docs/img/dataset/sysn_web_doc.png and b/docs/img/dataset/sysn_web_doc.png differ diff --git a/docs/img/dataset/titel_set_question.png b/docs/img/dataset/titel_set_question.png index c19d435..cb27006 100644 Binary files a/docs/img/dataset/titel_set_question.png and b/docs/img/dataset/titel_set_question.png differ diff --git a/docs/img/dataset/upload_web_doc.png b/docs/img/dataset/upload_web_doc.png index 04af539..543e959 100644 Binary files a/docs/img/dataset/upload_web_doc.png and b/docs/img/dataset/upload_web_doc.png differ diff --git a/docs/img/dataset/view_edit.png b/docs/img/dataset/view_edit.png index 20cdbba..941e376 100644 Binary files a/docs/img/dataset/view_edit.png and b/docs/img/dataset/view_edit.png differ diff --git a/docs/img/dataset/web_ sync.png b/docs/img/dataset/web_ sync.png deleted file mode 100644 index c8f911b..0000000 Binary files a/docs/img/dataset/web_ sync.png and /dev/null differ diff --git a/docs/img/dataset/web_dataset.png b/docs/img/dataset/web_dataset.png index 4e0bd32..547ca07 100644 Binary files a/docs/img/dataset/web_dataset.png and b/docs/img/dataset/web_dataset.png differ diff --git a/docs/img/dataset/web_doc.png b/docs/img/dataset/web_doc.png index ca23494..8a79cfa 100644 Binary files a/docs/img/dataset/web_doc.png and b/docs/img/dataset/web_doc.png differ diff --git a/docs/img/dataset/web_doc_content_area.png b/docs/img/dataset/web_doc_content_area.png new file mode 100644 index 0000000..57670b1 Binary files /dev/null and b/docs/img/dataset/web_doc_content_area.png differ diff --git a/docs/img/dataset/web_sync.png b/docs/img/dataset/web_sync.png new file mode 100644 index 0000000..68fbe7c Binary files /dev/null and b/docs/img/dataset/web_sync.png differ diff --git a/docs/img/dev/app_apikey.png b/docs/img/dev/app_apikey.png index 3864a54..92fa70a 100644 Binary files a/docs/img/dev/app_apikey.png and b/docs/img/dev/app_apikey.png differ diff --git a/docs/img/dev/app_chat.png b/docs/img/dev/app_chat.png index 53e60f4..6c29a6d 100644 Binary files a/docs/img/dev/app_chat.png and b/docs/img/dev/app_chat.png differ diff --git a/docs/img/dev/app_profile.png b/docs/img/dev/app_profile.png index a13e336..da2763f 100644 Binary files a/docs/img/dev/app_profile.png and b/docs/img/dev/app_profile.png differ diff --git a/docs/img/dev/app_swaagger.png b/docs/img/dev/app_swaagger.png deleted file mode 100644 index 9d60741..0000000 Binary files a/docs/img/dev/app_swaagger.png and /dev/null differ diff --git a/docs/img/dev/app_swagger.png b/docs/img/dev/app_swagger.png new file mode 100644 index 0000000..7e5eb3c Binary files /dev/null and b/docs/img/dev/app_swagger.png differ diff --git a/docs/img/dev/chat_open.png b/docs/img/dev/chat_open.png index e75d114..9751350 100644 Binary files a/docs/img/dev/chat_open.png and b/docs/img/dev/chat_open.png differ diff --git a/docs/img/dev/openai_baseurl.png b/docs/img/dev/openai_baseurl.png index 3d80650..7e5eb3c 100644 Binary files a/docs/img/dev/openai_baseurl.png and b/docs/img/dev/openai_baseurl.png differ diff --git a/docs/img/fx/add_fx.png b/docs/img/fx/add_fx.png index 74ce6bd..7bbecb3 100644 Binary files a/docs/img/fx/add_fx.png and b/docs/img/fx/add_fx.png differ diff --git a/docs/img/fx/copy_fx.jpg b/docs/img/fx/copy_fx.jpg index 47db985..9b94fc7 100644 Binary files a/docs/img/fx/copy_fx.jpg and b/docs/img/fx/copy_fx.jpg differ diff --git a/docs/img/fx/del_fx.png b/docs/img/fx/del_fx.png index 7e0d996..b085d2a 100644 Binary files a/docs/img/fx/del_fx.png and b/docs/img/fx/del_fx.png differ diff --git a/docs/img/fx/fx_debug.png b/docs/img/fx/fx_debug.png index ada842e..72f0ce5 100644 Binary files a/docs/img/fx/fx_debug.png and b/docs/img/fx/fx_debug.png differ diff --git a/docs/img/fx/use_del_fx.png b/docs/img/fx/use_del_fx.png index 194bb0d..824ee6a 100644 Binary files a/docs/img/fx/use_del_fx.png and b/docs/img/fx/use_del_fx.png differ diff --git a/docs/img/fx/use_fx.png b/docs/img/fx/use_fx.png index b630711..8f10337 100644 Binary files a/docs/img/fx/use_fx.png and b/docs/img/fx/use_fx.png differ diff --git a/docs/img/index/1panel.jpg b/docs/img/index/1panel.jpg index da2f2b5..1465ed8 100644 Binary files a/docs/img/index/1panel.jpg and b/docs/img/index/1panel.jpg differ diff --git a/docs/img/index/1panel_maxkb.jpg b/docs/img/index/1panel_maxkb.jpg index d93b9d8..bae245b 100644 Binary files a/docs/img/index/1panel_maxkb.jpg and b/docs/img/index/1panel_maxkb.jpg differ diff --git a/docs/img/index/UI.jpg b/docs/img/index/UI.jpg index 624b52b..93149be 100644 Binary files a/docs/img/index/UI.jpg and b/docs/img/index/UI.jpg differ diff --git a/docs/img/index/login.jpg b/docs/img/index/login.jpg index 3b643ff..b4bc71a 100644 Binary files a/docs/img/index/login.jpg and b/docs/img/index/login.jpg differ diff --git a/docs/img/index/maxkb-start.jpg b/docs/img/index/maxkb-start.jpg index 9f3cb80..78dafa2 100644 Binary files a/docs/img/index/maxkb-start.jpg and b/docs/img/index/maxkb-start.jpg differ diff --git a/docs/img/index/maxkb_setting.jpg b/docs/img/index/maxkb_setting.jpg index f1f7288..33c7768 100644 Binary files a/docs/img/index/maxkb_setting.jpg and b/docs/img/index/maxkb_setting.jpg differ diff --git a/docs/img/model/AWS_LLM.png b/docs/img/model/AWS_LLM.png index c0c171a..ffac2db 100644 Binary files a/docs/img/model/AWS_LLM.png and b/docs/img/model/AWS_LLM.png differ diff --git a/docs/img/model/aliyun_bailian_apikey.png b/docs/img/model/aliyun_bailian_apikey.png deleted file mode 100644 index e3620a2..0000000 Binary files a/docs/img/model/aliyun_bailian_apikey.png and /dev/null differ diff --git a/docs/img/model/aliyun_bailian_model.png b/docs/img/model/aliyun_bailian_model.png deleted file mode 100644 index f8318bc..0000000 Binary files a/docs/img/model/aliyun_bailian_model.png and /dev/null differ diff --git a/docs/img/model/aws_embed.png b/docs/img/model/aws_embed.png index f89d308..f6d3ec8 100644 Binary files a/docs/img/model/aws_embed.png and b/docs/img/model/aws_embed.png differ diff --git a/docs/img/model/azure_model.png b/docs/img/model/azure_model.png index d78c601..ebc419f 100644 Binary files a/docs/img/model/azure_model.png and b/docs/img/model/azure_model.png differ diff --git a/docs/img/model/bailian_asr.png b/docs/img/model/bailian_asr.png deleted file mode 100644 index 7f28f09..0000000 Binary files a/docs/img/model/bailian_asr.png and /dev/null differ diff --git a/docs/img/model/bailian_embed.png b/docs/img/model/bailian_embed.png deleted file mode 100644 index 872db57..0000000 Binary files a/docs/img/model/bailian_embed.png and /dev/null differ diff --git a/docs/img/model/bailian_llm.png b/docs/img/model/bailian_llm.png deleted file mode 100644 index ef44f62..0000000 Binary files a/docs/img/model/bailian_llm.png and /dev/null differ diff --git a/docs/img/model/bailian_reranker.png b/docs/img/model/bailian_reranker.png deleted file mode 100644 index 554a68b..0000000 Binary files a/docs/img/model/bailian_reranker.png and /dev/null differ diff --git a/docs/img/model/bailian_tts.png b/docs/img/model/bailian_tts.png deleted file mode 100644 index 4b30888..0000000 Binary files a/docs/img/model/bailian_tts.png and /dev/null differ diff --git a/docs/img/model/bailian_vision.png b/docs/img/model/bailian_vision.png deleted file mode 100644 index fbc7aba..0000000 Binary files a/docs/img/model/bailian_vision.png and /dev/null differ diff --git a/docs/img/model/bailian_vision_gen1.png b/docs/img/model/bailian_vision_gen1.png deleted file mode 100644 index 3a2ece3..0000000 Binary files a/docs/img/model/bailian_vision_gen1.png and /dev/null differ diff --git a/docs/img/model/bailian_vision_gen2.png b/docs/img/model/bailian_vision_gen2.png deleted file mode 100644 index b0b697b..0000000 Binary files a/docs/img/model/bailian_vision_gen2.png and /dev/null differ diff --git a/docs/img/model/doubao_APIkey.jpg b/docs/img/model/doubao_APIkey.jpg deleted file mode 100644 index 7ea28fb..0000000 Binary files a/docs/img/model/doubao_APIkey.jpg and /dev/null differ diff --git a/docs/img/model/doubao_LLM.jpg b/docs/img/model/doubao_LLM.jpg deleted file mode 100644 index 782d1ba..0000000 Binary files a/docs/img/model/doubao_LLM.jpg and /dev/null differ diff --git a/docs/img/model/doubao_gen1.png b/docs/img/model/doubao_gen1.png deleted file mode 100644 index 79db15b..0000000 Binary files a/docs/img/model/doubao_gen1.png and /dev/null differ diff --git a/docs/img/model/doubao_gen2.png b/docs/img/model/doubao_gen2.png deleted file mode 100644 index a073919..0000000 Binary files a/docs/img/model/doubao_gen2.png and /dev/null differ diff --git a/docs/img/model/doubao_jieru.jpg b/docs/img/model/doubao_jieru.jpg deleted file mode 100644 index 7e76db4..0000000 Binary files a/docs/img/model/doubao_jieru.jpg and /dev/null differ diff --git a/docs/img/model/doubao_jieru.png b/docs/img/model/doubao_jieru.png deleted file mode 100644 index d328fb8..0000000 Binary files a/docs/img/model/doubao_jieru.png and /dev/null differ diff --git a/docs/img/model/doubao_llm.png b/docs/img/model/doubao_llm.png deleted file mode 100644 index 39ff65a..0000000 Binary files a/docs/img/model/doubao_llm.png and /dev/null differ diff --git a/docs/img/model/doubao_llm_apikey.png b/docs/img/model/doubao_llm_apikey.png deleted file mode 100644 index 82b27b5..0000000 Binary files a/docs/img/model/doubao_llm_apikey.png and /dev/null differ diff --git a/docs/img/model/doubao_stt.png b/docs/img/model/doubao_stt.png deleted file mode 100644 index 32fd44e..0000000 Binary files a/docs/img/model/doubao_stt.png and /dev/null differ diff --git a/docs/img/model/doubao_stt_model.png b/docs/img/model/doubao_stt_model.png deleted file mode 100644 index 85863b4..0000000 Binary files a/docs/img/model/doubao_stt_model.png and /dev/null differ diff --git a/docs/img/model/doubao_tts.png b/docs/img/model/doubao_tts.png deleted file mode 100644 index 550f8ff..0000000 Binary files a/docs/img/model/doubao_tts.png and /dev/null differ diff --git a/docs/img/model/doubao_tts_model.png b/docs/img/model/doubao_tts_model.png deleted file mode 100644 index a346271..0000000 Binary files a/docs/img/model/doubao_tts_model.png and /dev/null differ diff --git a/docs/img/model/doubao_vision.png b/docs/img/model/doubao_vision.png deleted file mode 100644 index cd5e33e..0000000 Binary files a/docs/img/model/doubao_vision.png and /dev/null differ diff --git a/docs/img/model/gemini_llm.png b/docs/img/model/gemini_llm.png index 8659109..3a7a2a9 100644 Binary files a/docs/img/model/gemini_llm.png and b/docs/img/model/gemini_llm.png differ diff --git a/docs/img/model/hunyuan_LLM.jpg b/docs/img/model/hunyuan_LLM.jpg deleted file mode 100644 index b4147fe..0000000 Binary files a/docs/img/model/hunyuan_LLM.jpg and /dev/null differ diff --git a/docs/img/model/hunyuan_LLM.png b/docs/img/model/hunyuan_LLM.png deleted file mode 100644 index c466272..0000000 Binary files a/docs/img/model/hunyuan_LLM.png and /dev/null differ diff --git a/docs/img/model/hunyuan_embed.png b/docs/img/model/hunyuan_embed.png deleted file mode 100644 index 81e241c..0000000 Binary files a/docs/img/model/hunyuan_embed.png and /dev/null differ diff --git a/docs/img/model/hunyuan_vision.png b/docs/img/model/hunyuan_vision.png deleted file mode 100644 index 1f6a30b..0000000 Binary files a/docs/img/model/hunyuan_vision.png and /dev/null differ diff --git a/docs/img/model/hunyuan_vision_gen1.png b/docs/img/model/hunyuan_vision_gen1.png deleted file mode 100644 index 4606e8e..0000000 Binary files a/docs/img/model/hunyuan_vision_gen1.png and /dev/null differ diff --git a/docs/img/model/hunyuan_vision_gen2.png b/docs/img/model/hunyuan_vision_gen2.png deleted file mode 100644 index 2b1fd3a..0000000 Binary files a/docs/img/model/hunyuan_vision_gen2.png and /dev/null differ diff --git a/docs/img/model/kimi_apikey.png b/docs/img/model/kimi_apikey.png deleted file mode 100644 index 324ec93..0000000 Binary files a/docs/img/model/kimi_apikey.png and /dev/null differ diff --git a/docs/img/model/kimi_llm.png b/docs/img/model/kimi_llm.png deleted file mode 100644 index 2922c66..0000000 Binary files a/docs/img/model/kimi_llm.png and /dev/null differ diff --git a/docs/img/model/kimi_model.png b/docs/img/model/kimi_model.png deleted file mode 100644 index 80ff1b4..0000000 Binary files a/docs/img/model/kimi_model.png and /dev/null differ diff --git a/docs/img/model/local_embed.png b/docs/img/model/local_embed.png index 63a226f..9d2c12b 100644 Binary files a/docs/img/model/local_embed.png and b/docs/img/model/local_embed.png differ diff --git a/docs/img/model/local_reranker.png b/docs/img/model/local_reranker.png index 43ab6e2..9ace8ef 100644 Binary files a/docs/img/model/local_reranker.png and b/docs/img/model/local_reranker.png differ diff --git a/docs/img/model/model_add_para.png b/docs/img/model/model_add_para.png index 7880c67..7f55b58 100644 Binary files a/docs/img/model/model_add_para.png and b/docs/img/model/model_add_para.png differ diff --git a/docs/img/model/model_para_setting.png b/docs/img/model/model_para_setting.png index 73f7dcd..7b1274e 100644 Binary files a/docs/img/model/model_para_setting.png and b/docs/img/model/model_para_setting.png differ diff --git a/docs/img/model/ollama_embedding_model.png b/docs/img/model/ollama_embedding_model.png index 188bf80..b4df593 100644 Binary files a/docs/img/model/ollama_embedding_model.png and b/docs/img/model/ollama_embedding_model.png differ diff --git a/docs/img/model/ollama_model.png b/docs/img/model/ollama_model.png index e44254f..5c3f053 100644 Binary files a/docs/img/model/ollama_model.png and b/docs/img/model/ollama_model.png differ diff --git a/docs/img/model/ollama_vision.png b/docs/img/model/ollama_vision.png index 2ce99a9..00d0c70 100644 Binary files a/docs/img/model/ollama_vision.png and b/docs/img/model/ollama_vision.png differ diff --git a/docs/img/model/openai_asr.png b/docs/img/model/openai_asr.png index 73b9b1e..77a1782 100644 Binary files a/docs/img/model/openai_asr.png and b/docs/img/model/openai_asr.png differ diff --git a/docs/img/model/openai_embed.png b/docs/img/model/openai_embed.png index 1e2e197..2b964a3 100644 Binary files a/docs/img/model/openai_embed.png and b/docs/img/model/openai_embed.png differ diff --git a/docs/img/model/openai_llm.png b/docs/img/model/openai_llm.png index 6fdcf44..32ff714 100644 Binary files a/docs/img/model/openai_llm.png and b/docs/img/model/openai_llm.png differ diff --git a/docs/img/model/openai_tts.png b/docs/img/model/openai_tts.png index f4ad1da..5e645c1 100644 Binary files a/docs/img/model/openai_tts.png and b/docs/img/model/openai_tts.png differ diff --git a/docs/img/model/openai_vision.png b/docs/img/model/openai_vision.png index 08a4151..5ee711b 100644 Binary files a/docs/img/model/openai_vision.png and b/docs/img/model/openai_vision.png differ diff --git a/docs/img/model/qianfan-app.png b/docs/img/model/qianfan-app.png deleted file mode 100644 index 7577257..0000000 Binary files a/docs/img/model/qianfan-app.png and /dev/null differ diff --git a/docs/img/model/qianfan_app.png b/docs/img/model/qianfan_app.png deleted file mode 100644 index bb377b6..0000000 Binary files a/docs/img/model/qianfan_app.png and /dev/null differ diff --git a/docs/img/model/qianfan_embedding.png b/docs/img/model/qianfan_embedding.png deleted file mode 100644 index 7ffee01..0000000 Binary files a/docs/img/model/qianfan_embedding.png and /dev/null differ diff --git a/docs/img/model/qianfan_model.jpg b/docs/img/model/qianfan_model.jpg deleted file mode 100644 index b6a7b74..0000000 Binary files a/docs/img/model/qianfan_model.jpg and /dev/null differ diff --git a/docs/img/model/qianfan_model.png b/docs/img/model/qianfan_model.png deleted file mode 100644 index a274aed..0000000 Binary files a/docs/img/model/qianfan_model.png and /dev/null differ diff --git a/docs/img/model/tencent_apikey.png b/docs/img/model/tencent_apikey.png deleted file mode 100644 index 63bb2b4..0000000 Binary files a/docs/img/model/tencent_apikey.png and /dev/null differ diff --git a/docs/img/model/tongyi_apikey.png b/docs/img/model/tongyi_apikey.png deleted file mode 100644 index 51fe817..0000000 Binary files a/docs/img/model/tongyi_apikey.png and /dev/null differ diff --git a/docs/img/model/tongyi_llm.png b/docs/img/model/tongyi_llm.png deleted file mode 100644 index 13193cb..0000000 Binary files a/docs/img/model/tongyi_llm.png and /dev/null differ diff --git a/docs/img/model/tongyi_model.png b/docs/img/model/tongyi_model.png deleted file mode 100644 index 3267dcc..0000000 Binary files a/docs/img/model/tongyi_model.png and /dev/null differ diff --git a/docs/img/model/tongyi_vision.png b/docs/img/model/tongyi_vision.png deleted file mode 100644 index c918681..0000000 Binary files a/docs/img/model/tongyi_vision.png and /dev/null differ diff --git a/docs/img/model/tongyi_vision_gen1.png b/docs/img/model/tongyi_vision_gen1.png deleted file mode 100644 index 0e3113c..0000000 Binary files a/docs/img/model/tongyi_vision_gen1.png and /dev/null differ diff --git a/docs/img/model/tongyi_vision_gen2.png b/docs/img/model/tongyi_vision_gen2.png deleted file mode 100644 index 7072eb2..0000000 Binary files a/docs/img/model/tongyi_vision_gen2.png and /dev/null differ diff --git a/docs/img/model/vLLM_llm.png b/docs/img/model/vLLM_llm.png index c363bf3..2c0da27 100644 Binary files a/docs/img/model/vLLM_llm.png and b/docs/img/model/vLLM_llm.png differ diff --git a/docs/img/model/xinfo_asr.png b/docs/img/model/xinfo_asr.png index 2dc5231..a847ac3 100644 Binary files a/docs/img/model/xinfo_asr.png and b/docs/img/model/xinfo_asr.png differ diff --git a/docs/img/model/xinfo_embed.png b/docs/img/model/xinfo_embed.png index 63dd90a..fb243fd 100644 Binary files a/docs/img/model/xinfo_embed.png and b/docs/img/model/xinfo_embed.png differ diff --git a/docs/img/model/xinfo_llm.png b/docs/img/model/xinfo_llm.png index 5a03411..4241939 100644 Binary files a/docs/img/model/xinfo_llm.png and b/docs/img/model/xinfo_llm.png differ diff --git a/docs/img/model/xinfo_reranker.png b/docs/img/model/xinfo_reranker.png index 30489f6..d92c753 100644 Binary files a/docs/img/model/xinfo_reranker.png and b/docs/img/model/xinfo_reranker.png differ diff --git a/docs/img/model/xinfo_tts.png b/docs/img/model/xinfo_tts.png index 2a2a4bc..8176209 100644 Binary files a/docs/img/model/xinfo_tts.png and b/docs/img/model/xinfo_tts.png differ diff --git a/docs/img/model/xunfei_api.png b/docs/img/model/xunfei_api.png deleted file mode 100644 index 2ee5fb4..0000000 Binary files a/docs/img/model/xunfei_api.png and /dev/null differ diff --git a/docs/img/model/xunfei_app.png b/docs/img/model/xunfei_app.png deleted file mode 100644 index c32c5de..0000000 Binary files a/docs/img/model/xunfei_app.png and /dev/null differ diff --git a/docs/img/model/xunfei_embed.png b/docs/img/model/xunfei_embed.png deleted file mode 100644 index 1d85861..0000000 Binary files a/docs/img/model/xunfei_embed.png and /dev/null differ diff --git a/docs/img/model/xunfei_iat.png b/docs/img/model/xunfei_iat.png deleted file mode 100644 index abe71b3..0000000 Binary files a/docs/img/model/xunfei_iat.png and /dev/null differ diff --git a/docs/img/model/xunfei_iat_wss.png b/docs/img/model/xunfei_iat_wss.png deleted file mode 100644 index 53d1106..0000000 Binary files a/docs/img/model/xunfei_iat_wss.png and /dev/null differ diff --git a/docs/img/model/xunfei_llm.png b/docs/img/model/xunfei_llm.png deleted file mode 100644 index fc9b8c7..0000000 Binary files a/docs/img/model/xunfei_llm.png and /dev/null differ diff --git a/docs/img/model/zhipu_apikey.png b/docs/img/model/zhipu_apikey.png deleted file mode 100644 index 4a3ab9e..0000000 Binary files a/docs/img/model/zhipu_apikey.png and /dev/null differ diff --git a/docs/img/model/zhipu_llm.png b/docs/img/model/zhipu_llm.png deleted file mode 100644 index 02e9145..0000000 Binary files a/docs/img/model/zhipu_llm.png and /dev/null differ diff --git a/docs/img/model/zhipu_model.png b/docs/img/model/zhipu_model.png deleted file mode 100644 index d97ac1d..0000000 Binary files a/docs/img/model/zhipu_model.png and /dev/null differ diff --git a/docs/img/model/zhipu_vision.png b/docs/img/model/zhipu_vision.png deleted file mode 100644 index 742f1b1..0000000 Binary files a/docs/img/model/zhipu_vision.png and /dev/null differ diff --git a/docs/img/model/zhipu_vision_gen1.png b/docs/img/model/zhipu_vision_gen1.png deleted file mode 100644 index 2e34c76..0000000 Binary files a/docs/img/model/zhipu_vision_gen1.png and /dev/null differ diff --git a/docs/img/model/zhipu_vision_gen2.png b/docs/img/model/zhipu_vision_gen2.png deleted file mode 100644 index 5176f85..0000000 Binary files a/docs/img/model/zhipu_vision_gen2.png and /dev/null differ diff --git a/docs/img/system/CAS.png b/docs/img/system/CAS.png index 7f4ba34..e917ec2 100644 Binary files a/docs/img/system/CAS.png and b/docs/img/system/CAS.png differ diff --git a/docs/img/system/LDAP.png b/docs/img/system/LDAP.png index 23b2542..78db718 100644 Binary files a/docs/img/system/LDAP.png and b/docs/img/system/LDAP.png differ diff --git a/docs/img/system/OIDC.png b/docs/img/system/OIDC.png index 3e78252..a411e62 100644 Binary files a/docs/img/system/OIDC.png and b/docs/img/system/OIDC.png differ diff --git a/docs/img/system/add_user.png b/docs/img/system/add_user.png index fb98809..417ca25 100644 Binary files a/docs/img/system/add_user.png and b/docs/img/system/add_user.png differ diff --git a/docs/img/system/apikey.png b/docs/img/system/apikey.png index f15a61c..c1bddb9 100644 Binary files a/docs/img/system/apikey.png and b/docs/img/system/apikey.png differ diff --git a/docs/img/system/apikey_setting.png b/docs/img/system/apikey_setting.png index 2eec711..e9c8590 100644 Binary files a/docs/img/system/apikey_setting.png and b/docs/img/system/apikey_setting.png differ diff --git a/docs/img/system/auth_login.png b/docs/img/system/auth_login.png index 354b8b5..200f964 100644 Binary files a/docs/img/system/auth_login.png and b/docs/img/system/auth_login.png differ diff --git a/docs/img/system/oauth2.png b/docs/img/system/oauth2.png index 6c8cf6c..509cfc1 100644 Binary files a/docs/img/system/oauth2.png and b/docs/img/system/oauth2.png differ diff --git a/docs/img/system/swagger_api.png b/docs/img/system/swagger_api.png index f808cf2..92fa70a 100644 Binary files a/docs/img/system/swagger_api.png and b/docs/img/system/swagger_api.png differ diff --git a/docs/img/system/theme.png b/docs/img/system/theme.png index 7c962b9..d95aa3c 100644 Binary files a/docs/img/system/theme.png and b/docs/img/system/theme.png differ diff --git a/docs/img/system/user.png b/docs/img/system/user.png index 21feea1..ba25906 100644 Binary files a/docs/img/system/user.png and b/docs/img/system/user.png differ diff --git a/docs/img/team/addMembers.png b/docs/img/team/addMembers.png index ef494d3..7453341 100644 Binary files a/docs/img/team/addMembers.png and b/docs/img/team/addMembers.png differ diff --git a/docs/img/team/del_member.png b/docs/img/team/del_member.png index d4df4b6..acf4bb7 100644 Binary files a/docs/img/team/del_member.png and b/docs/img/team/del_member.png differ diff --git a/docs/img/team/permission_setting.png b/docs/img/team/permission_setting.png index 9e4bdb3..d879933 100644 Binary files a/docs/img/team/permission_setting.png and b/docs/img/team/permission_setting.png differ diff --git a/docs/img/team/teamlist.png b/docs/img/team/teamlist.png index 54c4b1c..4aa37d4 100644 Binary files a/docs/img/team/teamlist.png and b/docs/img/team/teamlist.png differ diff --git a/docs/installation/aliyun.md b/docs/installation/aliyun.md deleted file mode 100644 index b728aac..0000000 --- a/docs/installation/aliyun.md +++ /dev/null @@ -1,44 +0,0 @@ -# Alibaba Cloud Marketplace Deployment Guide - -!!! Abstract "" - - This guide will explain how to purchase, deploy and use MaxKB images through Alibaba Cloud Marketplace, and provide discounted links for purchasing servers. - -## 1 Purchase Image - -!!! Abstract "" - - - MaxKB is available on Alibaba Cloud Marketplace. You can purchase the image directly through this link: - [MaxKB Cloud Marketplace Purchase Link](https://market.aliyun.com/products/53690006/cmjj00067609.html?userCode=kmemb8jp) - - - You can also purchase an Alibaba Cloud server separately and search for **MaxKB** when selecting an image for quick deployment. - - **Note:** Please select an appropriate instance specification when purchasing, recommended 4 cores/8GB RAM or higher. - -!!! Abstract "Server Discounts" - - If you don't have a server yet, you can purchase Alibaba Cloud servers through this discount link: - - - [Exclusive Alibaba Cloud Special Price Link - 45% Off](https://market.aliyun.com/common/dashi/maxkb?userCode=kmemb8jp) - -## 2 Start Service - -!!! Abstract "" - - After the image starts, you can access MaxKB through your browser at: - - ``` - Service Access Address: http://SERVER-IP:8080 - Default Username: admin - Default Password: MaxKB@123.. - ``` - - After first login, it is recommended to change the password immediately and configure other security settings. - -## 3 Open Ports - -!!! Abstract "" - - - To ensure external access to MaxKB services, you need to open port `8080` in your Alibaba Cloud server's security group rules. - - - For detailed steps on opening ports, please refer to Alibaba Cloud's [Port Opening Tutorial](https://help.aliyun.com/document_detail/25471.html). diff --git a/docs/installation/offline_installtion.md b/docs/installation/offline_installtion.md deleted file mode 100644 index 51e4c84..0000000 --- a/docs/installation/offline_installtion.md +++ /dev/null @@ -1,116 +0,0 @@ -## 1 Deployment Requirements - -!!! Abstract "" - For production environments, offline installation is the officially recommended. - -### 1.1 Server Configuration - -!!! Abstract "" - - **Server Requirements:** - - * Operating System: Ubuntu 22.04 / CentOS 7.6 64-bit system - * CPU/Memory: 4 cores/8GB or higher - * Disk Space: 100GB - -### 1.2 Port Requirements - -!!! Abstract "" - The following ports should be accessible: - -| Port | Purpose | Description | -|---------|:----------|:----------------------------------------------------------| -| 22 | SSH | Used for installation, upgrades and management | -| 8080 | Web Port | Default web service access port, can be changed as needed | - -## 2 Download Installation Package - -!!! Abstract "" - Open the [FIT2CLOUD Open Source Community MaxKB Community Edition Downloads](https://community.fit2cloud.com/#/products/maxkb/downloads) page to download the latest version installation package and upload it to the deployment server (using v1.2.0 as an example). - -## 3 Installation and Deployment - -### 3.1 Extract Installation Package - -!!! Abstract "" - - Log in to the deployment server as root user via SSH protocol and extract the installation package: - ``` - tar -zxvf maxkb-v1.2.0-offline.tar.gz - ``` - -### 3.2 Installation Configuration (Optional) - -!!! Abstract "" - - MaxKB installation directory, service running port, database configuration and other parameter can be configured in the install.conf file within the installation package. - - ``` - ## Installation directory - MAXKB_BASE=/opt - ## Service port - MAXKB_PORT=8080 - ## Docker subnet settings - MAXKB_DOCKER_SUBNET=172.19.0.0/16 - # Database configuration - ## Whether to use external database - MAXKB_EXTERNAL_PGSQL=false - ## Database host - MAXKB_PGSQL_HOST=pgsql - ## Database port - MAXKB_PGSQL_PORT=5432 - ## Database name - MAXKB_PGSQL_DB=maxkb - ## Database username - MAXKB_PGSQL_USER=root - ## Database password - MAXKB_PGSQL_PASSWORD=Password123@postgres - ``` - - **Note**: Parameters can be modified in the install.conf file before first installation, and the installation will be executed according to the modified parameters. If you need to modify configuration parameters after installation is complete, you need to make changes in the ${MAXKB_BASE}/maxkb/.env file and execute the `mkctl reload` command to reload the configuration file. - -### 3.3 Execute Installation Script - -!!! Abstract "" - - ``` - # Enter the extracted installation package directory - cd maxkb-v1.2.0-offline - - # Execute installation Script - bash install.sh - ``` - -![Installation](../img/index/install.jpg) - -## 4 Login Access - -!!! Abstract "" - - Once all container statuses show as `healthy`, you can access the address `http://IP:8080` through a browser and log in to MaxKB using the default administrator credentials. - - ``` - Username: admin - Password: MaxKB@123.. - ``` - -![Login](../img/index/login.jpg) - -## 5 Offline Upgrade - -!!! Abstract "" - - Following the above instructions, download and upload the new version installation package, extract and execute the installation command again to upgrade. - - ``` - # Enter new version directory - cd maxkb-v1.x.x-offline - - # Run installation script - /bin/bash install.sh - - # Check MaxKB running status - mkctl status - ``` - - **Note:** Please backup the database before upgrading. diff --git a/docs/quick_start.md b/docs/quick_start.md index 33c052c..2c9ad19 100644 --- a/docs/quick_start.md +++ b/docs/quick_start.md @@ -1,8 +1,3 @@ ---- -hide: - - navigation ---- - # Quick Start ## Installing MaxKB diff --git a/docs/stylesheets/extra.css b/docs/stylesheets/extra.css index d82bb61..f13ae65 100644 --- a/docs/stylesheets/extra.css +++ b/docs/stylesheets/extra.css @@ -1,7 +1,12 @@ -:root > * { - --md-primary-fg-color: #5A55FA; - --md-primary-fg-color--light: #5A55FA; - --md-primary-fg-color--dark: #5A55FA; +:root > * { + --md-primary-fg-color: #5a55fa; + --md-primary-fg-color--light: #5a55fa; + --md-primary-fg-color--dark: #5a55fa; +} + +.twemoji { + display: inline-block !important; + box-shadow: none !important; } /* 不添加额外样式的图片样式 */ @@ -14,7 +19,7 @@ /* 不添加浏览器导航栏的图片样式 */ .md-content img:not(.browser-mockup img) { display: block; - box-shadow: 0 0.1em .5em 0 rgba(0, 0, 0, 0.4); + box-shadow: 0 0.1em 0.5em 0 rgba(0, 0, 0, 0.4); border: none; border-radius: 4px; } @@ -35,28 +40,31 @@ /* 添加浏览器导航栏的图片样式 */ .browser-mockup { border-top: 2em solid rgba(230, 230, 230, 0.8); - box-shadow: 0 0.1em .5em 0 rgba(0, 0, 0, 0.4); + box-shadow: 0 0.1em 0.5em 0 rgba(0, 0, 0, 0.4); position: relative; - border-radius: 4px 4px 0 0 + border-radius: 4px 4px 0 0; } .browser-mockup:before { display: block; position: absolute; - content: ''; + content: ""; top: -1.25em; left: 1em; width: 0.5em; height: 0.5em; border-radius: 50%; background-color: #f44; - box-shadow: 0 0 0 2px #f44, 1.5em 0 0 2px #9b3, 3em 0 0 2px #fb5; + box-shadow: + 0 0 0 2px #f44, + 1.5em 0 0 2px #9b3, + 3em 0 0 2px #fb5; } .browser-mockup.with-tab:after { display: block; position: absolute; - content: ''; + content: ""; top: -2em; left: 5.5em; width: 20%; @@ -69,7 +77,7 @@ .browser-mockup.with-url:after { display: block; position: absolute; - content: ''; + content: ""; top: -1.6em; left: 5.5em; width: calc(100% - 6em); @@ -78,10 +86,10 @@ background-color: white; } -.browser-mockup>* { +.browser-mockup > * { display: block; } -.browser-mockup>a { +.browser-mockup > a { line-height: 0; -} \ No newline at end of file +} diff --git a/docs/user_manual/X-Pack/app_integrate.md b/docs/user_manual/X-Pack/app_integrate.md deleted file mode 100644 index c6574a6..0000000 --- a/docs/user_manual/X-Pack/app_integrate.md +++ /dev/null @@ -1,201 +0,0 @@ -# 应用接入 - - -!!! Abstract "" - MaxKB X-Pack 支持创建的应用的与企业微信应用、企业微信客服、公众号（服务号和订阅号）、钉钉应用、飞书应用接入，实现企业内部员工、外部公众进行对话。 - -### 1 企业微信应用接入 - -!!! Abstract "" - （1）在应用接入中点击【企业微信应用】的【配置】按钮，配置信息中会自动生成回调 URL,并需要复制到企业微信应用的 API 接收设置中，除此之外的其它信息，将在企业微信管理后台中生成并获取。 - -![显示设置](../../img/app/ent_wechat_setting.png) - -!!! Abstract "" - （2）获取企业ID: 在[企业微信管理后台](https://work.weixin.qq.com/wework_admin/frame#apps)，点击 【我的企业】菜单，在最下方可以看到企业ID 信息。 - -![企业ID](../../img/app/enterprise_id.png) - -!!! Abstract "" - （3）创建企业微信应用：在【应用管理】中点击【创建应用】，选择应用logo，设置应用名称和应用介绍以及可见范围。 - -![创建企业微信应用](../../img/app/create_robot.png) -![企业微信应用设置](../../img/app/robot_info.png) - -!!! Abstract "" - （4）获取 Agentid 和 Secret:进入创建的应用，获取 AgentId 和 Secret。点击【查看】Secret 将通过企业微信进行查看。 - -![Agentid](../../img/app/Agentid.png) - -!!! Abstract "" - （5）获取 Token 和 EncodingAESKey：进入企业微信应用设置的【接收消息】-【设置API接收】，随机获取 Token 和EncodingAESKey。然后在 MaxKB 企业微信应用配置输入对应的参数并保存，然后再回到企业微信应用设置【接收消息】-【设置API接收】企业微信应用接入配置中自动生成的回调 URL，然后再点击【保存】。 - - **注意：** 在企业微信应用API接收消息配置保存之前，一定要在 MaxKB 中完成企业微信应用配置并保存，不然保存时将报错`openapi回调地址请求不通过`。 - -![设置API接收](../../img/app/recmsg_api.png) - - -!!! Abstract "" - **注意：** 在企业微信应用API接收消息配置保存之前，一定要在 MaxKB 中完成企业微信应用配置并保存，不然保存时将报错`openapi回调地址请求不通过`。 - -![设置API接收](../../img/app/recmsg_api_setting.png) - - - -!!! Abstract "" - （6）配置企业可信IP：进入企业微信应用设置的【企业可信IP】中，把 MaxKB 服务的IP地址设置为可信IP。 - -![企业微信对话](../../img/app/creditable_ip.png) - -!!! Abstract "" - - 以上步骤配置完成后，并在企业微信应用接入配置中输入企业微信应用信息后保存，便可在企业微信应用中找到机器人进行对话了。 - -![企业微信对话](../../img/app/enter_wechat_dialog.png) - -### 2 企业微信客服对接 - -!!! Abstract "" - MaxKB 应用接入企业微信应用后，在此基础上，还可以将此应用接入到微信客服，对外部用户提供服务。 - （1）创建客服账号：在企业微信管理后台的【应用管理】中，点击【微信客服】，并创建账号,客服账号的接待方式选择：机器人+人工接待 - -![企业微信客服账号](../../img/app/weichat_customservice.png){width="800px"} - -![企业微信客服账号](../../img/app/create_cs_account.png){width="800px"} - -!!! Abstract "" - （2）客服账号与应用关联：在微信客服中配置【可调用接口的应用】，并配置客服账号。 - -![企业微信客服账号](../../img/app/cs_app.png){width="800px"} - -![企业微信客服账号](../../img/app/config_cs_account.png){width="800px"} - -!!! Abstract "" - （3）开启客服助理：在企业的外部群设置中开启【客服助理】，微信客服便会加入群中，群里的任何用户仅需要@客服助理，即可发起咨询，并微信客服进行一对一的对话。 - -![企业微信客服账号](../../img/app/enable_cs.png){width="500px"} - -![企业微信客服账号](../../img/app/dialog_cs.png){width="500px"} - -### 3 公众号接入 - -!!! Abstract "" - （1）在应用接入中点击【公众号】的【配置】按钮，配置信息中会自动生成回调 URL,并需要复制到【微信公众平台-设置与开发-基本配置-服务器配置】服务器地址URL 中，除此之外的其它信息，将在微信公众平台中生成并获取。 -![公众号配置](../../img/app/public_account_setting_empty.png) - -!!! Abstract "" - （2）获取APP ID、APP Secret、Token和消息加解密密钥：在微信公众平台-设置与开发-基本配置中获取APP ID、APP Secret信息，并启动服务器配置，生成 Token 和消息加解密密钥。 -![公众号信息](../../img/app/public_account_basicinfo.png) -![公众号信息](../../img/app/public_account_setting1.png) - -!!! Abstract "" - （3）将生成的APP ID、APP Secret、Token 和消息加解密密钥信息输入到MaxKB 公众号配置窗口中，并保存。 -![公众号信息](../../img/app/public_account_setting2.png) - -!!! Abstract "" - （4）回到第（2）步，将【公众号-基本配置-服务器配置】中的 URL 设置为 MaxKB 公众号配置窗口中的回调地址 URL，然后提交，并启用服务器配置。 -![公众号信息](../../img/app/public_account_setting_url.png) - -!!! Abstract "" - （5）开启客服接口权限。 -![客服接口权限](../../img/app/public_account_enable_api.png) - -!!! Abstract "" - （6）设置IP白名单：进入公众号基本配置的【IP白名单】中，添加 MaxKB 服务器的 IP 地址。 -![公众号IP白名单](../../img/app/public_account_ip.png) - -!!! Abstract "" - 根据以上步骤完成配置后，即可在公众号中发送消息进行对话了。 - -![公众号对话](../../img/app/public_account_dialog.jpeg){width="500px"} - - -### 4 钉钉应用接入 - -!!! Abstract "" - （1）在应用接入中点击【钉钉应用】的【配置】按钮，配置信息中会自动生成回调 URL,并需要复制到钉钉-机器人设置-消息接收地址中，其它信息，将在钉钉开放平台中生成并获取。 -![钉钉机器人创建](../../img/app/dingding_setting.png) - -!!! Abstract "" - （2）创建机器人：在 [钉钉开放平台](https://open-dev.dingtalk.com/)的【应用开发】-【钉钉应用】中，点击【创建应用】，填写应用信息后，点击【保存】。 -![钉钉机器人创建](../../img/app/dingding_robot_create.png) - -!!! Abstract "" - （3）获取Client ID和Client Secret： 进入创建好的钉钉应用，打开【凭证与基础信息】页面，获取Client ID和Client Secret。 -![钉钉应用信息](../../img/app/dingding_info.png) - -!!! Abstract "" - （4）在MaxKB 钉钉应用配置窗口中输入步骤获取的Client ID和Client Secret 应用凭证，并保存。 -![钉钉应用信息](../../img/app/dingding_setting2.png) - -!!! Abstract "" - （5）点击【添加应用能力】，选择 【机器人】能力，点击 【添加】，输入机器人基本配置信息后，将【消息接收模式】设置为HTTP模式，并将 MaxKB 钉钉应用配置中的回调地址的URL填写到消息接收地址中，然后点击【发布】。 -![钉钉机器人添加](../../img/app/dingding_robot_add.png) -![钉钉机器人息接收模式](../../img/app/dingding_msg_recvmode.png) - -!!! Abstract "" - 根据以上步骤完成配置后，就可以群里@机器人中进行对话，或与机器人应用一对一对话。 -![钉钉对话](../../img/app/dingding_dialog.jpeg){width="500px"} - -### 5 飞书应用接入 - -!!! Abstract "" - （1）在应用接入中点击【飞书应用】的【配置】按钮，配置信息中会自动生成回调 URL，并需要复制到【飞书开放平台-事件与回调-事件配置-配置订阅方式】的请求地址中，除此之外的其它信息，将在飞书开放平台中生成并获取。 - -![飞书应用配置](../../img/app/feishu_setting.png) - -!!! Abstract "" - （2）创建飞书机器人：首先在 [飞书开发平台](https://open.feishu.cn/app/)的【企业自建应用】中，点击【创建企业自建应用】，填写应用信息。 -![飞书添加机器人](../../img/app/feishu_create_app.png) - -!!! Abstract "" - 点击【添加应用能力】，选择 【按能力添加-机器人】，点击【添加】 -![飞书添加机器人](../../img/app/feishu_add_robot.png) - -!!! Abstract "" - （3）添加权限：打开【权限管理】，复制下面的权限配置并粘贴到【API权限-权限配置】输入框，全选筛选出来的权限项，点击【批量开通】，最后点击【确认】。 - - ``` - m:message,im:message.group_at_msg:readonly,im:message.p2p_msg:readonly,im:message:send_as_bot,im:resource, contact:contact.base:readonly，contact:user.base:readonly - ``` - -![飞书权限](../../img/app/feishu_app_auth.png) - -!!! Abstract "" - （4）获取APP ID、APP Secret和Verification Token：进入创建好的飞书应用，打开【凭证与基础信息】页面，获取APP ID和APP Secret。 -![飞书应用信息i](../../img/app/feishu_app_info.png) - - -!!! Abstract "" - 打开【事件与回调-加密策略】，获取 Verification Token。 - -![飞书token](../../img/app/feishu_verify_token.png) - -!!! Abstract "" - （5）在MaxKB 飞书应用配置窗口输入上述信息，并保存。 - -![飞书事件与回调](../../img/app/maxkb_feishu_config.png) - - -!!! Abstract "" - （6）配置回调URL：在飞书开放平台中打开【事件与回调】-【回调配置】的订阅方式中，将MaxKB中飞书应用的回调地址的URL填写到【请求地址】中。 -![飞书事件与回调](../../img/app/feishu_event_config1.png) - -![飞书事件与回调](../../img/app/feishu_event_config2.png) - - -!!! Abstract "" - （7）发布应用：点击【创建版本】，填写版本信息后，点击【保存】发。 -![飞书应用发布](../../img/app/feishu_app_create1.png) - -![飞书应用发布](../../img/app/feishu_app_create2.png) - -!!! Abstract "" - 确认发布后，应用状态将更新为`已启用`。 -![飞书应用发布](../../img/app/feishu_app_enabled.png) - -!!! Abstract "" - 根据以上步骤完成配置后，即可打开飞书客户端搜索`MaxKB小助手`，点击【添加】后进行一对一对话。 -![飞书对话](../../img/app/feishu_app_search.jpg){width="500px"} - -![飞书对话](../../img/app/feishu_dialog.jpeg){width="500px"} diff --git a/docs/user_manual/X-Pack/login_auth.md b/docs/user_manual/X-Pack/login_auth.md index b1f55db..377b715 100644 --- a/docs/user_manual/X-Pack/login_auth.md +++ b/docs/user_manual/X-Pack/login_auth.md @@ -1,19 +1,19 @@ # Login Authentication -!!! Abstract "" +!!! Abstract "" MaxKB Professional Edition supports LDAP, CAS, OIDC, OAUTH2 single sign-on protocols, as well as WeChat Work, DingTalk, and Lark mobile QR code login verification, meeting enterprise requirements for strong identity verification and access control. ## 1 LDAP -!!! Abstract "" - The LDAP configuration process can be referenced in the image below. Note that this feature is enabled by checking "Enable LDAP Authentication" at the bottom. - Tip: After configuration, you can click "Test Connection" at the top to instantly test if the configuration information is correct and the network is connected. +!!! Abstract "" + The LDAP configuration process can be referenced in the image below. Note that this feature is enabled by checking "Enable LDAP Authentication" at the bottom. + Tip: After configuration, you can click "Test Connection" at the top to instantly test if the configuration information is correct and the network is connected. ![LDAP](../../img/system/LDAP.png) ## 2 CAS !!! Abstract "" - The CAS configuration process can be referenced in the image below. + The CAS configuration process can be referenced in the image below. **Note:** The CAS callback URL is the MaxKB access URL plus `/api/cas`, for example: `http://40.100.86.240:8080/api/cas` ![CAS](../../img/system/CAS.png) @@ -21,7 +21,7 @@ ## 3 OIDC !!! Abstract "" - The OIDC configuration process can be referenced in the image below. + The OIDC configuration process can be referenced in the image below. **Note:** The OIDC callback URL is the MaxKB access URL plus `/api/oidc`, for example: `http://40.100.86.240:8080/api/oidc` ![OIDC](../../img/system/OIDC.png) @@ -32,96 +32,8 @@ **Note:** The OAUTH2 callback URL is the MaxKB access URL plus `/api/oauth2`, for example: `http://40.100.86.240:8080/api/oauth2` ![OAHTU2](../../img/system/oauth2.png) -## 5 QR Code - -### 5.1 WeChat Work QR Code Login - -!!! Abstract "" - WeChat Work QR code login configuration parameter description: - - - Corp ID: Enterprise ID, obtained by WeChat Work admin in the "Enterprise Information" section of the admin backend. - - Agent ID: Unique identifier for WeChat Work custom application, obtained when creating or viewing applications in the admin backend. - - App Secret: Secret key for WeChat Work custom application. - - Callback URL: The MaxKB access URL. - -!!! Abstract "" - To configure WeChat Work QR code login for MaxKB, the WeChat Work admin needs to log in to [WeChat Work Admin Portal](https://work.weixin.qq.com/wework_admin/frame) and create a custom enterprise application, then publish it after configuration. - Step 1: Create application. Go to [Application Management-Applications-Custom], click "Create Application", enter application name and other information. After creation, you can view the application's Agent ID and Secret. - -![WeChat Work App Creation](../../img/system/qiwei_create_app.png){width="500px"} - -![WeChat Work App Info](../../img/system/qiwei_appinfo.png){width="500px"} - -!!! Abstract "" - Step 2: Set trusted domain. Click "Set trusted domain name" in "Developer API ", add and verify trusted domain. -![Set Trusted Domain](../../img/system/qiwei_yuming.png){width="500px"} - -!!! Abstract "" - Step 3: Set authorization callback domain. Configure the callback domain in "Log in to via authorization by WeCom". -![Set Callback Domain](../../img/system/qiwei_callback.png){width="500px"} - -!!! Abstract "" - Step 3: Configure enterprise trusted IPs. Configure trusted IPs in "Company's Trusted IP". -![Configure Trusted IPs](../../img/system/qiwei_ip.png){width="500px"} - -!!! Abstract "" - After completing WeChat Work application configuration and publishing, configure corresponding information in MaxKB WeChat Work login QR code configuration page and verify. -![WeChat Work Configuration](../../img/system/qiwei_setting.png) - -### 5.2 DingTalk QR Code Login - -!!! Abstract "" - DingTalk QR code login configuration parameter description: - - - Corp ID: DingTalk organization identifier, Shown in organization information at top right of DingTalk Open Platform. - - APP Key: DingTalk application identifier, Shown in application "Credentials and Basic Information". - - App Secret: DingTalk application secret, Shown in application "Credentials and Basic Information". - - Callback URL: The MaxKB access URL. - -!!! Abstract "" - To configure DingTalk QR code login for MaxKB, you need to create and configure an application in [DingTalk Open Platform](https://open-dev.dingtalk.com/). - - Step 1: Create application. Go to [Application Development-DingTalk Applications], click "Create Application". After creation, you can view APPKey and APPSecret in "Credentials and Basic Information". -![DingTalk Create App](../../img/system/dingding_app_create.png) -![DingTalk App Info](../../img/system/dingding_app_info.png) - -!!! Abstract "" - Step 2: Publish application. In "Version Management and Publishing", fill in application version number, version description and other information, click "Save" to publish the application. -![DingTalk App Info](../../img/system/dingding_app_release.png) - -!!! Abstract "" - After completing configuration and publishing, configure in MaxKB DingTalk QR code login configuration page and save. -![DingTalk Create App](../../img/system/dingding_setting.png) - -### 5.3 Lark QR Code Login - -!!! Abstract "" - Feishu QR code login configuration parameter description: - - - APP Key: Application key - - Corp ID: Organization ID - - App Secret: Application secret - - Callback URL: MaxKB URL + '/api/feishu', for example: `http://40.100.86.243/api/feishu` - -![Feishu Configuration](../../img/system/feishu_setting.png) - -!!! Abstract "" - To configure Feishu QR code login for MaxKB, you need to create an enterprise custom application in [Feishu Open Platform](https://open.feishu.cn/), complete configuration and publish, then configure in MaxKB Feishu QR code login configuration page and save. - -!!! Abstract "" - Step 1: Create enterprise custom application. Click "Create Enterprise Custom Application", enter application name, description and upload application icon, then click "Create". -![Feishu Create](../../img/system/feishu_create_app.png) - -!!! Abstract "" - Step 2: Configure redirect URL. In "Development Configuration-Security Settings", enter the callback URL from MaxKB Feishu QR code configuration popup, click "Add". -![Feishu Configure Callback URL](../../img/system/feishu_url.png) - -!!! Abstract "" - Step 3: Publish application. In "Version Management and Publishing", enter application version number, update notes and other information, click "Save" to complete application publishing. -![Feishu Publish App](../../img/system/feishu_app_release.png) - ## 6 Login Authentication !!! Abstract "" - After completing authentication configuration, select the corresponding authentication method on the login page to log in. + After completing authentication configuration, select the corresponding authentication method on the login page to log in. ![Login Authentication](../../img/system/auth_login.png) diff --git a/docs/user_manual/app/app-view.md b/docs/user_manual/app/app-view.md index f543aca..a51bcbf 100644 --- a/docs/user_manual/app/app-view.md +++ b/docs/user_manual/app/app-view.md @@ -2,41 +2,41 @@ # Application Overview !!! Abstract "" - On the application list page, click an application panel to enter the application overview page. + On the application list page, click an application panel to enter the application overview page. On the overview page, you can enable/disable/regenerate public access links, demo the application, embed in third-party systems, set access restrictions and manage API Keys. ![Application Overview](../../img/index/UI.jpg) -## 1 Public Access Link +## 1 Public URL !!! Abstract "" - Click Demo or copy the public access link to a browser to enter the Q&A page. + Click **Preview** or copy the Public URL to a browser to enter the Q&A page. ![Q&A](../../img/app/QA.png) !!! Abstract "" - Public access links support toggling and regeneration. - For active public links, if you disable the current public link or regenerate a new public access link, visitors will see the message: Sorry, the service is currently under maintenance and unavailable. Please try again later! + Public URL support toggling and regeneration. + For active public URLs, if you disable the current public URL or regenerate a new public URL, visitors will see the message: Sorry, the service is currently under maintenance and unavailable. Please try again later! ![Service Closed](../../img/app/app-closed.png) -## 2 Third-party Embedding +## 2 Get Embed Code !!! Abstract "" - MaxKB applications support zero-code embedding into third-party web system. - On the application overview page, click Embed in website to copy either fullscreen mode code or floating mode code for embedding into third-party web system. Once embedded, you can interact with MaxKB within the third-party system. + MaxKB applications support zero-code embedding into third-party web system. + On the application overview page, click **Get Embed Code** to copy either fullscreen mode code or floating mode code for embedding into third-party web system. Once embedded, you can interact with MaxKB within the third-party system. -![Embed in Third-party](<../../img/app/embed.png>) +![Get Embed Code](<../../img/app/embed.png>) ## 3 Access Control !!! Abstract "" MaxKB supports setting question limits per client and embedding whitelist hotlink protection for applications, as well as configuring whether to display knowledge sources to end users. -![Access Restrictions](<../../img/app/Access restrictions.png>) +![Access Restrictions](<../../img/app/access_restrictions.png>) ## 4 API Access Credentials !!! Abstract "" If external systems need to interact with MaxKB through API calls, you'll need to check the API documentation and the application's API Key. Click API Key button to open the API Key management dialog, where you can create, enable/disable, and delete API Keys. -![API key](../../img/app/app_apikey.png) \ No newline at end of file +![API key](../../img/app/app_apikey.png) diff --git a/docs/user_manual/dataset/dataset.md b/docs/user_manual/dataset/dataset.md index aded69f..783577c 100644 --- a/docs/user_manual/dataset/dataset.md +++ b/docs/user_manual/dataset/dataset.md @@ -5,14 +5,14 @@ * **General Knowledge Base**: For offline document, supporting text files, tables and Q&A pairs. * **Web Site Knowledge Base**: For online static text data, automatically synchronizing text data from root URL and sub-URLs after inputting the root web address. - + MaxKB supports knowledge base creation, re-vectorization, settings, synchronization, export, deletion, etc. ## 1 Create Knowledge Base !!! Abstract "" Open the Knowledge page, click Create Knowledge. - + ![dataset](../../img/dataset/create.jpg) ### 1.1 General Knowledge Base @@ -27,7 +27,7 @@ * Maximum 50 files per upload; * Each file not exceeding 100 MB; * Support selecting folders to upload files meeting format requirements. - + !!! Abstract "" **Note:** ZIP files include two forms: @@ -36,7 +36,7 @@ ![上传文档](../../img/dataset/create_offline_dataset.png) - + !!! Abstract "" MaxKB currently supports two methods for document segmentation: Intelligent Segmentation and Advanced Segmentation. @@ -44,7 +44,7 @@ (1) Markdown file automatic segmentation rules
- * Segmentation based on hierarchical titles (up to 6 levels), with maximum 4096 characters per segment; + * Segmentation based on hierarchical titles (up to 6 levels), with maximum 4096 characters per segment; * When text paragraphs under the last level exceed the set segment length, the system will look for line breaks within the segment length to split. (2) HTML and DOCX file automatic segmentation rules @@ -60,29 +60,29 @@ ![Automatic Segmentation](<../../img/dataset/automatic_paragraphing.png>) !!! Abstract "" - Document specification recommendations: + Document specification recommendations: - * **Standardized segment marker**: Offline documents should have standardized segment markers, otherwise the split paragraphs will be irregular. + * **Standardized segment marker**: Offline documents should have standardized segment markers, otherwise the split paragraphs will be irregular. * **Complete paragraphs**: A segment should ideally describe a complete piece of content. !!! Abstract "" - **Advanced Segmentation** + **Advanced Segmentation** Users can customize segment delimiters, segment length and automatic cleaning based on document specifications. - * Supported segment markers: #, ##, ###, ####, #####, ######, -, blank line, line break, space, semicolon, comma, period. Also supports manual input of other markers or regular expressions. - * Segment length: Length of a single segment, ranging from 50 to 4096 characters. - * Automatic cleaning: When enabled, the system automatically removes redundant symbols like spaces, blank lines, tabs, etc. + * Supported segment markers: #, ##, ###, ####, #####, ######, -, blank line, line break, space, semicolon, comma, period. Also supports manual input of other markers or regular expressions. + * Segment length: Length of a single segment, ranging from 50 to 4096 characters. + * Automatic cleaning: When enabled, the system automatically removes redundant symbols like spaces, blank lines, tabs, etc. ![Advanced Segmentation](<../../img/dataset/advanced_segmentation.png>) !!! Abstract "" - **Add "Releated Question" section for question-based QA pair during import** + **Add "Releated Question" section for question-based QA pair during import** When checked, all segment titles will be set as related questions for the segments. ![Set Title as Related Question](../../img/dataset/titel_set_question.png) !!! Abstract "" - **Preview** + **Preview** After setting segmentation rules, click [Generate Preview] to check the segmentation effect of the latest rules. ![Segmentation Preview](<../../img/dataset/preview_segmentation.png>) @@ -90,7 +90,7 @@ You can edit unreasonable segments and delete meaningless segments in the segmentation preview. ![Edit Segments](../../img/dataset/view_edit.png) - + !!! Abstract "" After clicking [Start Import], the system backend will automatically perform segmentation, storage, vectorization operations on the documents. When completed, the file status will show as complete. ![Edit Segments](../../img/dataset/processing.png) @@ -107,7 +107,7 @@ * Web Root URL: The root node URL of the online static website. The system will automatically fetch data from the root URL and its sub-level URLs. For example, using DataEase V2 online documentation, the Web root URL would be: https://dataease.io/docs/v2/ * Selector: You can set selectors to get data from specific div elements on the web page. By default, it gets data from the body. For example, to only get the middle content section of DataEase online documentation, you can right-click to inspect the webpage, locate the middle content area with div class "md-content", then enter the selector as: .md-content -![Alt text](../../img/dataset/DataEase_doc.png) +![Alt text](../../img/dataset/web_doc_content_area.png) !!! Abstract "" After clicking OK button, you'll be redirected to the document list page to view detailed information of documents under the current website. When processing is complete, the file status will show as complete. @@ -127,7 +127,7 @@ * Sync and Replace: Re-fetch web site documents and replace documents in the local knowledge base. * Full Sync: First delete all documents in the local knowledge base, then re-fetch web site documents. -![Knowledge Base Export](../../img/dataset/web_ sync.png) +![Knowledge Base Export](../../img/dataset/web_sync.png) ### 2.2 Vectorization @@ -159,4 +159,3 @@ **Note**: Knowledge base deletion cannot be undone, please proceed with caution. ![Knowledge Base Delete](../../img/dataset/dataset_del.png) - diff --git a/docs/user_manual/dataset/doclist.md b/docs/user_manual/dataset/doclist.md index 00b399f..3aa999e 100644 --- a/docs/user_manual/dataset/doclist.md +++ b/docs/user_manual/dataset/doclist.md @@ -11,7 +11,7 @@ ![Intelligent Segmentation](<../../img/dataset/auto_paragraph.png>) -!!! Abstract "" +!!! Abstract "" For Web site knowledge bases, click Import Document button to open a dialog box where you can enter document links and selectors. Multiple online Web documents can be entered line by line. ![Import Document](../../img/dataset/upload_web_doc.png) @@ -28,15 +28,15 @@ ## 3 Document Synchronization -!!! Abstract "" +!!! Abstract "" Web site knowledge bases support synchronization of selected documents. During synchronization, all segments under the current document will be deleted, and text data will be re-fetched from the document address and re-segmented. ![Sync Document](../../img/dataset/sysn_web_doc.png) ## 4 Hit Processing Settings -!!! Abstract "" - 文档设置目前支持文档的命中处理方式。 +!!! Abstract "" + Currently, the document settings support various ways of processing document hits: * Model Optimization: When segments under this document are matched during questioning, prompts will be generated according to the application's prompt templates and sent to the model for optimization before returning answers. * Direct Answer: When segments under this document are matched during questioning and similarity meets the threshold, segment content is returned directly. This method is recommended when images, links and other information need to be returned. @@ -45,35 +45,35 @@ ## 5 Generate Questions -!!! Abstract "" +!!! Abstract "" Select files and click Generate Questions button or execute generate questions operation. The AI model will summarize the file content to generate corresponding questions and automatically associate them. ![Generate Questions](../../img/dataset/gen_question.png) -## 6 Document Migration +## 6 Moving documents -!!! Abstract "" - Select documents and click [Migrate] button to migrate documents to other knowledge bases. +!!! Abstract "" + Select documents and click [Move] button to move documents to other knowledge bases. -![Migrate Document](../../img/dataset/move_web_doc.png) +![Moving documents](../../img/dataset/move_web_doc.png) ## 7 Export EXCEL/ZIP -!!! Abstract "" +!!! Abstract "" Select documents and execute Export Excel/Export Zip operation to download documents to local client. ![Document Export](../../img/dataset/dataset_file_export.png) ## 8 Delete Document -!!! Abstract "" +!!! Abstract "" Select documents and click delete button or execute delete operation to delete selected documents. ![Delete Document](../../img/dataset/doc_delete.png) ## 9 Enable and Disable Documents -!!! Abstract "" +!!! Abstract "" In the status column of the document list, you can enable or disable documents. When a document is disabled, the system will not search segments under that document when users ask questions. The system will only search after re-enabling. ![Enable Document](../../img/dataset/doc_enable.png) @@ -87,28 +87,28 @@ ### 10.1 Add Segment -!!! Abstract "" - Click Add Segment button to open the add segment dialog box. Fill in segment title, segment content (supports markdown style editing) and related questions, then click Submit button to add a new segment. +!!! Abstract "" + Click Add Segment button to open the add segment dialog box. Fill in segment title, segment content (supports markdown style editing) and related questions, then click Submit button to add a new segment. **Recommendation:** To accurately match segments, it's recommended to set related questions for segments. This will prioritize matching related questions before mapping segment content, thereby improving matching efficiency and accuracy. ![Add Segment](../../img/dataset/add_segmentation.png) ### 10.2 Edit Segment -!!! Abstract "" +!!! Abstract "" Click the segment panel to edit segment information and related questions on the segment details page. ![Segment Details](../../img/dataset/edit_segmentation.png) -### 10.3 Migrate Segment +### 10.3 Moving section -!!! Abstract "" - In the segment panel, you can migrate selected segments to documents in other knowledge bases. +!!! Abstract "" + In the segment panel, you can move selected segments to documents in other knowledge bases. -![Migrate Segment](../../img/dataset/move_segmentation.png) +![Moving section](../../img/dataset/move_segmentation.png) ### 10.4 Delete Segment -!!! Abstract "" +!!! Abstract "" In the segment panel, you can delete selected segments. ![Delete Segment](../../img/dataset/del_segmentation.png) diff --git a/docs/user_manual/dataset/problem.md b/docs/user_manual/dataset/problem.md index 011d837..8391d1b 100644 --- a/docs/user_manual/dataset/problem.md +++ b/docs/user_manual/dataset/problem.md @@ -1,24 +1,24 @@ -# 问题 +# Questions -!!! Abstract "" +!!! Abstract "" MaxKB knowledge base administrators can collect potential user questions, create questions and associate them with content in the knowledge base documents, maintain a list of frequently asked questions in the knowledge base to improve Q&A accuracy. ![Question List](../../img/dataset/problem_list.png) ## 1 Create Question -!!! Abstract "" +!!! Abstract "" Click Create Question button and enter questions line by line. ![Create Question](../../img/dataset/create_problem.png) ## 2 Associate Questions with Segments -!!! Abstract "" +!!! Abstract "" After adding questions, they can be associated with segments in documents. When users ask questions, the system will first match against the question bank to query related segments. ![Associate Questions with Segments](../../img/dataset/problem_segmentation.png) ## 3 Question Details -!!! Abstract "" +!!! Abstract "" Click a question to view details, where you can modify segment content or remove associations. -![Question Details](../../img/dataset/problem_detail.png) \ No newline at end of file +![Question Details](../../img/dataset/problem_detail.png) diff --git a/docs/user_manual/model/X-Infer_model.md b/docs/user_manual/model/X-Infer_model.md index 60e8377..e6c229a 100644 --- a/docs/user_manual/model/X-Infer_model.md +++ b/docs/user_manual/model/X-Infer_model.md @@ -1,37 +1,37 @@ -## 1 添加模型 +## 1 Add Model !!! Abstract "" - 选择模型供应商为`Xorbits Inference`，并在模型添加对话框中输入如下必要信息： + Select the model provider as `Xorbits Inference`, and enter the following necessary information in the model addition dialog box: - * 模型名称：MaxKB 中自定义的模型名称。 - * 权限：分为私有和公用两种权限，私有模型仅当前用户可用，公用模型即系统内所有用户均可使用，但其它用户不能编辑和删除。 - * 模型类型：大语言模型/向量模型/重排模型/语音识别/语音合成/图片理解/图片生成。 - * 基础模型：不同类型模型下的基础模型名称，下拉选项是常用的一些基础模型名称，支持自定义输入。 + * Model Name: The custom model name in MaxKB. + * Permissions: Divided into private and public permissions. Private models are available only to the current user, while public models can be used by all users within the system, but other users cannot edit or delete them. + * Model Type: Large Language Model / Vector Model / Rerank Model / Speech Recognition / Speech Synthesis / Image Understanding / Image Generation. + * Base Model: The base model name under different model types. The dropdown options include some commonly used base model names, and custom input is supported. -!!! Abstract "" - 大语言模型和重排模型需要输入 API 域名和API Key，向量模型需要输入API 域名。 +!!! Abstract "" + Large language models and rerank models require input of the API Domain and API Key, while vector models require input of the API Domain. - * API 域名：Xorbits Inference 服务地址，例如：http://192.168.20.242:9997 。 - * API Key：若没有 API Key，输入任意字符即可。 + * API Domain: Xorbits Inference service address, e.g., http://192.168.20.242:9997. + * API Key: If there is no API Key, you can enter any characters. -## 2 配置样例 +## 2 Configuration Example !!! Abstract "" - Xorbits Inference-大语言模型配置样例图示如下： + The configuration example illustration for Xorbits Inference-Large Language Model is as follows: ![Xorbits Inference](../../img/model/xinfo_llm.png){ width="500px" } !!! Abstract "" - Xorbits Inference-向量模型配置样例图示如下： + The configuration example illustration for Xorbits Inference-Vector Model is as follows: ![Xorbits Inference](../../img/model/xinfo_embed.png){ width="500px" } !!! Abstract "" - Xorbits Inference-重排模型配置样例图示如下： + The configuration example illustration for Xorbits Inference-Rerank Model is as follows: ![Xorbits Inference](../../img/model/xinfo_reranker.png){ width="500px" } !!! Abstract "" - Xorbits Inference-语音识别模型配置样例图示如下： + The configuration example illustration for Xorbits Inference-Speech Recognition Model is as follows: ![Xorbits Inference](../../img/model/xinfo_asr.png){ width="500px" } !!! Abstract "" - Xorbits Inference-语音合成模型配置样例图示如下： + The configuration example illustration for Xorbits Inference-Speech Synthesis Model is as follows: ![Xorbits Inference](../../img/model/xinfo_tts.png){ width="500px" } diff --git a/docs/user_manual/model/amazon_bedrock_model.md b/docs/user_manual/model/amazon_bedrock_model.md index 72d121f..408b208 100644 --- a/docs/user_manual/model/amazon_bedrock_model.md +++ b/docs/user_manual/model/amazon_bedrock_model.md @@ -1,23 +1,23 @@ -## 1 添加模型 +## 1 Adding a Model !!! Abstract "" - 选择模型供应商为`Amazon Bedrock`，并在模型添加对话框中输入如下必要信息： + Select the model provider as `Amazon Bedrock` and enter the following necessary information in the model addition dialog: - * 模型名称：MaxKB 中自定义的模型名称。 - * 权限：分为私有和公用两种权限，私有模型仅当前用户可用，公用模型即系统内所有用户均可使用，但其它用户不能编辑和删除。 - * 模型类型：大语言模型/向量大模型。 - * 基础模型：Amazon Bedrock 支持的 LLM 模型名称，下拉选项是常用的一些大语言模型名称，支持自定义输入。 - * Region Name：模型开通的区域。 - * Access Key ID/Secret Access Key: Access Key ID 和 Secret Access Key 是用于对 AWS 服务（包括 Amazon Bedrock）的编程访问进行身份验证的凭证。 + * Model Name: Custom model name in MaxKB. + * Permission: Divided into private and public permissions. Private models are only available to the current user, while public models can be used by all users within the system, but other users cannot edit or delete them. + * Model Type: Large language model/Vector large model. + * Base Model: The names of LLM models supported by Amazon Bedrock. The drop-down option includes commonly used large language model names and supports custom input. + * Region Name: The region where the model is activated. + * Access Key ID/Secret Access Key: The Access Key ID and Secret Access Key are credentials used to authenticate programmatic access to AWS services, including Amazon Bedrock. -## 2 配置样例 +## 2 Configuration Example !!! Abstract "" - Amazon Bedrock-大语言模型配置样例图示： + Example of Amazon Bedrock - Large Language Model configuration: -![AWS LLM模型](../../img/model/AWS_LLM.png){ width="500px" } +![AWS LLM Model](../../img/model/AWS_LLM.png){ width="500px" } !!! Abstract "" - Amazon Bedrock-向量模型配置样例图示： + Example of Amazon Bedrock - Vector Model configuration: -![AWS LLM模型](../../img/model/aws_embed.png){ width="500px" } \ No newline at end of file +![AWS Vector Model](../../img/model/aws_embed.png){ width="500px" } diff --git a/docs/user_manual/model/azure_openai_model.md b/docs/user_manual/model/azure_openai_model.md index 9d5b0ae..684226b 100644 --- a/docs/user_manual/model/azure_openai_model.md +++ b/docs/user_manual/model/azure_openai_model.md @@ -1,26 +1,26 @@ -## 1 添加模型 +## 1 Adding a Model -!!! Abstract "" - 添加 Azure OpenAI 大模型之前，需要先在 [Azure AI Studio](https://ai.azure.com/) 中注册，并获取有关API 域名、API Key、部署详细信息等内容，参考下图： +!!! Abstract "" + Before adding an Azure OpenAI large model, you need to register in [Azure AI Studio](https://ai.azure.com/) and obtain information regarding the API domain, API Key, deployment details, etc. Refer to the image below: ![Azure OpenAI Key](<../../img/model/Azure_APIKey.png>) -![Azure OpenAI DemployInfo](<../../img/model/Azure_deployInfo.png>) +![Azure OpenAI Deployment Info](<../../img/model/Azure_deployInfo.png>) -!!! Abstract "" - 选择模型供应商为`Azure OpenAI`，并在模型添加对话框中输入如下必要信息： +!!! Abstract "" + Select the model provider as `Azure OpenAI` and enter the following necessary information in the model addition dialog: - * 模型名称：MaxKB 中自定义的模型名称。 - * 权限：分为私有和公用两种权限，私有模型仅当前用户可用，公用模型即系统内所有用户均可使用，但其它用户不能编辑和删除。 - * 模型类型：大语言模型/向量模型/语音识别/语音合成/图片理解/图片生成。 - * 基础模型：具体的基础模型由部署名决定，见上图。 - * API 版本：模型版本 - * API 域名：Azure OpenAI 项目 API 服务 URL，见上图。 - * API Key：Azure OpenAI 项目 API 服务认证验证信息，见上图。 - * 部署名：Azure AI Studio 项目操场中模型的部署名称。 + * Model Name: Custom model name in MaxKB. + * Permission: Divided into private and public permissions. Private models are only available to the current user, while public models can be used by all users within the system, but other users cannot edit or delete them. + * Model Type: Large language model/Vector model/Speech recognition/Speech synthesis/Image understanding/Image generation. + * Base Model: The specific base model is determined by the deployment name, as shown in the above image. + * API Version: Model version. + * API Domain: The URL of the API service for the Azure OpenAI project, as shown in the above image. + * API Key: The authentication information for the Azure OpenAI project API service, as shown in the above image. + * Deployment Name: The deployment name of the model in Azure AI Studio's project playground. -## 2 配置样例 +## 2 Configuration Example -!!! Abstract "" - Azure OpenAI-大语言模型配置样例图示： +!!! Abstract "" + Example of Azure OpenAI - Large Language Model configuration: -![Azure_大语言模型](../../img/model/azure_model.png) \ No newline at end of file +![Azure Large Language Model](../../img/model/azure_model.png) diff --git a/docs/user_manual/model/bailian_model.md b/docs/user_manual/model/bailian_model.md deleted file mode 100644 index df15d18..0000000 --- a/docs/user_manual/model/bailian_model.md +++ /dev/null @@ -1,51 +0,0 @@ -## 1 添加模型 - -!!! Abstract "" - 选择模型供应商为`阿里云百炼`，并在模型添加对话框中输入如下必要信息： - - * 模型名称：MaxKB 中自定义的模型名称。 - * 权限：分为私有和公用两种权限，私有模型仅当前用户可用，公用模型即系统内所有用户均可使用，但其它用户不能编辑和删除。 - * 模型类型：大语言模型/向量模型/重排模型/语音识别/语音合成/图片理解/图片生成。 - * 基础模型：不同类型模型下的基础模型名称，下拉选项是常用的一些基础模型名称，支持自定义输入。 - * API 域名：模型服务 API 服务访问地址，目前当模型类型是大语言模型时需要输入。 - * API Key：模型服务 API 服务访问密钥。 - -![阿里云百炼 APIKEY](../../img/model/aliyun_bailian_apikey.png) - -## 2 配置样例 - -!!! Abstract "" - 阿里云百炼-大语言模型配置样例图示如下，需要说明的是：不同的大语言模型对应的 API 域名不一样，具体请查看对应基础模型的 API调用示例。 - -![阿里云百炼 大语言模型配置](../../img/model/bailian_llm.png){ width="500px" } - - -!!! Abstract "" - 阿里云百炼-向量模型配置样例图示如下： - -![阿里云百炼 向量模型配置](../../img/model/bailian_embed.png){ width="500px" } - -!!! Abstract "" - 阿里云百炼-重排模型配置样例图示如下： -![阿里云百炼 重排模型配置](../../img/model/bailian_reranker.png){ width="500px" } - -!!! Abstract "" - 阿里云百炼-语音识别模型配置样例图示如下： -![阿里云百炼 语音识别模型配置](../../img/model/bailian_asr.png){ width="500px" } - -!!! Abstract "" - 阿里云百炼-语音合成模型配置样例图示如下： -![阿里云百炼 语音合成模型配置](../../img/model/bailian_tts.png){ width="500px" } - -!!! Abstract "" - 阿里云百炼-图片理解模型配置样例图示如下： -![阿里云百炼 图片理解模型配置](../../img/model/bailian_vision.png){ width="500px" } - -!!! Abstract "" - 阿里云百炼-图片生成模型默认图像尺寸为 1024 * 1024，图片数量 1 张，风格为 ，即由模型随机输出图像风格，配置样例图示如下： -![阿里云百炼 图片生成模型配置](../../img/model/bailian_vision_gen1.png){ width="500px" } - -![阿里云百炼 图片生成模型配置](../../img/model/bailian_vision_gen2.png){ width="500px" } - - - diff --git a/docs/user_manual/model/deepseek_model.md b/docs/user_manual/model/deepseek_model.md deleted file mode 100644 index 6edb4c5..0000000 --- a/docs/user_manual/model/deepseek_model.md +++ /dev/null @@ -1,21 +0,0 @@ -## 1 添加模型 - -!!! Abstract "" - 添加 DeepSeek 大模型之前，需要先在 [DeepSeek 开放平台](https://platform.deepseek.com/) 创建 API Key。 -![kimi 模型](../../img/model/deepseek_apikey.png) - -!!! Abstract "" - 选择模型供应商为`DeepSeek`，并在模型添加对话框中输入如下必要信息： - - * 模型名称：MaxKB 中自定义的模型名称。 - * 权限：分为私有和公用两种权限，私有模型仅当前用户可用，公用模型即系统内所有用户均可使用，但其它用户不能编辑和删除。 - * 模型类型：大语言模型。 - * 基础模型：不同类型模型下的基础模型名称，下拉选项是常用的一些基础模型名称，支持自定义输入。 - * API Key：在 DeepSeek 开放平台创建并查看。 - -## 2 配置样例 - -!!! Abstract "" - DeepSeek-大语言模型配置样例图示： - -![deepseek 模型](../../img/model/deepseek_llm.png){ width="500px" } \ No newline at end of file diff --git a/docs/user_manual/model/doubao_model.md b/docs/user_manual/model/doubao_model.md deleted file mode 100644 index 3f9f9c8..0000000 --- a/docs/user_manual/model/doubao_model.md +++ /dev/null @@ -1,55 +0,0 @@ -## 1 添加模型 - -!!! Abstract "" - 选择模型供应商为`火山引擎`，并在模型添加对话框中输入如下必要信息： - - * 模型名称：MaxKB 中自定义的模型名称。 - * 权限：分为私有和公用两种权限，私有模型仅当前用户可用，公用模型即系统内所有用户均可使用，但其它用户不能编辑和删除。 - * 模型类型：大语言模型/语音识别/语音合成/图片理解/图片生成。 - -!!! Abstract "" - 豆包大语言模型其它接入参数说明： - - * 基础模型：在线推理的接入点ID。 - * API 域名：https://ark.cn-beijing.volces.com/api/v3 - * API Key：接入点-> API调用 -> 通过API Key授权 -> 选择 API Key 复制获取，如下图 - -![火山引擎 LLM模型](../../img/model/doubao_jieru.png) -![火山引擎 LLM模型](../../img/model/doubao_llm_apikey.png) - - -!!! Abstract "" - 语音识别和语音合成需要输入App ID、Access Token、Cluster ID。 - - * App 域名：语音识别模型 API 域名为`wss://openspeech.bytedance.com/api/v2/asr`，语音合成模型 API 域名为`wss://openspeech.bytedance.com/api/v1/tts/ws_binary` - * App ID：在火山引擎中创建的应用ID。 - * Access Token：在火山引擎应用对应的认证信息。 - * Cluster ID：在火山引擎应用对应的服务信息. - -![火山引擎_语音识别](../../img/model/doubao_stt.png) -![火山引擎o_语音合成](../../img/model/doubao_tts.png) - -## 2 配置样例 - -!!! Abstract "" - 火山引擎-大语言模型配置样例图示如下： -![火山引擎 LLM模型](../../img/model/huoshan_llm.png){ width="500px" } - -!!! Abstract "" - 火山引擎-语音识别模型配置样例图示如下： -![火山引擎 LLM模型](../../img/model/huoshan_asr.png){ width="500px" } - -!!! Abstract "" - 火山引擎-语音合成模型配置样例图示如下： -![火山引擎 LLM模型](../../img/model/huoshan_tts.png){ width="500px" } - -!!! Abstract "" - 火山引擎-图片理解模型配置样例图示如下： -![火山引擎 图片理解](../../img/model/doubao_vision.png){ width="500px" } - -!!! Abstract "" - 火山引擎-图片生成模型需要在火山引擎【控制台-访问控制-API访问密钥】新建密钥，样例图示如下： -![火山引擎 图片生成](../../img/model/doubao_gen1.png){ width="500px" } - -![火山引擎 图片生成](../../img/model/doubao_gen2.png){ width="500px" } - diff --git a/docs/user_manual/model/gemini_model.md b/docs/user_manual/model/gemini_model.md index 9b6ce5f..04d3275 100644 --- a/docs/user_manual/model/gemini_model.md +++ b/docs/user_manual/model/gemini_model.md @@ -1,23 +1,23 @@ -## 1 添加模型 +## 1 Adding a Model !!! Abstract "" - 添加 Gemini 大模型之前，需要先在 [Google AI Studio](https://aistudio.google.com/) 创建 API Key。 -![Gemini APIKEy](../../img/model/gemini_key.png) + Before adding a Gemini large model, you need to create an API Key in [Google AI Studio](https://aistudio.google.com/). +![Gemini API Key](../../img/model/gemini_key.png) !!! Abstract "" - 选择模型供应商为`Gemini`，并在模型添加对话框中输入如下必要信息： + Select the model provider as `Gemini` and enter the following necessary information in the model addition dialog: - * 模型名称：MaxKB 中自定义的模型名称。 - * 权限：分为私有和公用两种权限，私有模型仅当前用户可用，公用模型即系统内所有用户均可使用，但其它用户不能编辑和删除。 - * 模型类型：大语言模型/向量模型/语音识别/图片理解。 - * 基础模型：不同类型模型下的基础模型名称，下拉选项是常用的一些基础模型名称，支持自定义输入。 - * API Key：获取 API Key。 + * Model Name: Custom model name in MaxKB. + * Permission: Divided into private and public permissions. Private models are only available to the current user, while public models can be used by all users within the system, but other users cannot edit or delete them. + * Model Type: Large language model/Vector model/Speech recognition/Image understanding. + * Base Model: The name of the base model under different types, with a drop-down option for some commonly used base model names, and supports custom input. + * API Key: Obtain the API Key. - **注意：** 使用 Gemini API 需要确保程序所在服务器位于 [Gemini API 所支持的地区](https://ai.google.dev/gemini-api/docs/available-regions?hl=zh-cn) ，否则无法调用API，并且无法进入Google AI Studio。 + **Note:** To use the Gemini API, ensure that the server where the program is located is in a [region supported by Gemini API](https://ai.google.dev/gemini-api/docs/available-regions?hl=en), otherwise you cannot call the API and access Google AI Studio. -## 2 配置样例 +## 2 Configuration Example !!! Abstract "" - Gemini-大语言模型配置样例图示： + Example of Gemini - Large Language Model configuration: -![gemini 大语言模型](../../img/model/gemini_llm.png){ width="500px" } \ No newline at end of file +![Gemini Large Language Model](../../img/model/gemini_llm.png){ width="500px" } diff --git a/docs/user_manual/model/hunyuan_model.md b/docs/user_manual/model/hunyuan_model.md deleted file mode 100644 index 0186de6..0000000 --- a/docs/user_manual/model/hunyuan_model.md +++ /dev/null @@ -1,38 +0,0 @@ -## 1 添加模型 - -!!! Abstract "" - 添加腾讯混元大模型之前，需要先在腾讯云开通腾讯混元服务并新建秘钥。 - -![tencent apikey](../../img/model/tencent_apikey.png) - -!!! Abstract "" - 选择模型供应商为`腾讯混元`，并在模型添加对话框中输入如下必要信息： - - * 模型名称：MaxKB 中自定义的模型名称。 - * 权限：分为私有和公用两种权限，私有模型仅当前用户可用，公用模型即系统内所有用户均可使用，但其它用户不能编辑和删除。 - * 模型类型：大语言模型/向量模型/图片理解/图片生成。 - * 基础模型：不同类型模型下的基础模型名称，下拉选项是常用的一些基础模型名称，支持自定义输入。 - * APPID：【腾讯云-控制台-访问管理-API密钥管理】中获取。 - * SecretId：【腾讯云-控制台-访问管理-API密钥管理】中获取。 - * SecretKey：【腾讯云-控制台-访问管理-API密钥管理】新建密钥时下载的SecretKey。 - -## 2 配置样例 - -!!! Abstract "" - 腾讯混元-大语言模型配置样例图示如下： - -![hunyuan 大语言模型](../../img/model/hunyuan_LLM.png){ width="500px" } - -!!! Abstract "" - 腾讯混元-向量模型配置样例图示如下： -![hunyuan 向量模型](../../img/model/hunyuan_embed.png){ width="500px" } - -!!! Abstract "" - 腾讯混元-图片理解模型配置样例图示如下： -![hunyuan 图片理解模型](../../img/model/hunyuan_vision.png){ width="500px" } - -!!! Abstract "" - 腾讯混元-图片生成模型默认图像尺寸为 768 * 768，图片数量 1 张，风格为201，即日系动漫风格，默认配置样例图示如下： -![hunyuan 图片生成模型](../../img/model/hunyuan_vision_gen1.png){ width="500px" } - -![hunyuan 图片生成模型](../../img/model/hunyuan_vision_gen2.png){ width="500px" } diff --git a/docs/user_manual/model/kimi_model.md b/docs/user_manual/model/kimi_model.md deleted file mode 100644 index 749a32d..0000000 --- a/docs/user_manual/model/kimi_model.md +++ /dev/null @@ -1,23 +0,0 @@ -## 1 添加模型 - -!!! Abstract "" - 添加 kimi 模型之前，需要先在 [Moonshot AI 开放平台](https://platform.moonshot.cn/console/account) 中注册并创建 API Key。 - -![kimi APIkey](../../img/model/kimi_apikey.png) - -!!! Abstract "" - 选择模型供应商为`Kimi`，并在模型添加对话框中输入如下必要信息： - - * 模型名称：MaxKB 中自定义的模型名称。 - * 权限：分为私有和公用两种权限，私有模型仅当前用户可用，公用模型即系统内所有用户均可使用，但其它用户不能编辑和删除。 - * 模型类型：大语言模型。 - * 基础模型：不同类型模型下的基础模型名称，下拉选项是常用的一些基础模型名称，支持自定义输入。。 - * API 域名：https://api.moonshot.cn/v1 - * API Key：在 Kimi 账户中心的 API Key 管理中创建和查看。 - -## 2 配置样例 - -!!! Abstract "" - kimi-大语言模型配置样例图示： - -![kimi 模型](../../img/model/kimi_llm.png){ width="500px" } \ No newline at end of file diff --git a/docs/user_manual/model/local_model.md b/docs/user_manual/model/local_model.md index 72c30e8..231fb18 100644 --- a/docs/user_manual/model/local_model.md +++ b/docs/user_manual/model/local_model.md @@ -1,36 +1,37 @@ -## 1 下载模型 +## 1 Download the Model !!! Abstract "" - **前提条件：** 将模型下载到服务器，并且挂在到 MaxKB 容器中,以向量模型为例详细说明如下。 + **Prerequisite:** Download the model to the server and mount it to the MaxKB container. Using a vector model as an example, the detailed explanation is as follows. - （1）将向量模型下载到本地服务器。 + (1) Download the vector model to the local server. ``` - # 推荐模型下载网址 + # Recommended model download website https://huggingface.co/models?other=text-embedding - # 下载后存放在/opt/maxkb/model/local_embedding目录下 + # After downloading, store it in the /opt/maxkb/model/local_embedding directory ``` - （2）使用-v 将宿主机模型路径挂载到 MaxKB 容器内部。 + (2) Use -v to mount the host model path inside the MaxKB container. ``` -v /opt/maxkb/model/local_embedding:/opt/maxkb/model/local_embedding - # 说明： v 模型所在目录:MaxKB容器内目录 + # Note: v Host model directory: MaxKB container internal directory ``` -## 2 添加模型 + +## 2 Add the Model !!! Abstract "" - 在模型管理中，点击供应商【本地模型】，直接进入下一步填写本地模型的表单。 + In the model management, click on the supplier [Local Model] to directly proceed to the next step and fill out the form for the local model. - * 模型名称：MaxKB 中自定义的模型名称。 - * 权限：分为私有和公用两种权限，私有模型仅当前用户可用，公用模型即系统内所有用户均可使用，但其它用户不能编辑和删除。 - * 模型类型：向量模型/重排模型。 - * 基础模型：模型在 MaxKB 容器下的绝对路径。 - * 模型目录：模型的目录（基础模型为名称时，该目录生效，基础模型若有绝对路径，则该参数不生效，建议与基础模型填写一致）。 + * Model Name: Custom model name in MaxKB. + * Permission: Divided into private and public permissions. Private models are only available to the current user, while public models can be used by all users within the system, but other users cannot edit or delete them. + * Model Type: Vector model/Reranker model. + * Base Model: The absolute path of the model in the MaxKB container. + * Model Directory: The directory of the model (this takes effect when the base model is a name; if the base model has an absolute path, this parameter does not take effect, and it's recommended to match it with the base model). -## 3 配置样例 +## 3 Configuration Example !!! Abstract "" - 本地模型-向量模型配置样例图示如下： -![本地向量模型](../../img/model/local_embed.png){ width="500px" } + Example of configuration for a local model - Vector Model: +![Local Vector Model](../../img/model/local_embed.png){ width="500px" } !!! Abstract "" - 本地模型-重排模型配置样例图示如下： -![本地重排模型](../../img/model/local_reranker.png){ width="500px" } \ No newline at end of file + Example of configuration for a local model - Reranker Model: +![Local Reranker Model](../../img/model/local_reranker.png){ width="500px" } diff --git a/docs/user_manual/model/model_param.md b/docs/user_manual/model/model_param.md index 1b9b765..ca70222 100644 --- a/docs/user_manual/model/model_param.md +++ b/docs/user_manual/model/model_param.md @@ -1,5 +1,4 @@ - !!! Abstract "" - 点击模型右上角更多菜单，点击【模型参数设置】，可以根据各平台提供的其它参数进行配置。 -![模型参数设置](../../img/model/model_para_setting.png) -![添加模型参数](../../img/model/model_add_para.png ) \ No newline at end of file + Click on the more menu at the top right corner of the model, and then click on [Model Parameter Settings] to configure other parameters provided by each platform. +![Model Parameter Settings](../../img/model/model_para_setting.png) +![Add Model Parameters](../../img/model/model_add_para.png) diff --git a/docs/user_manual/model/model_ summary.md b/docs/user_manual/model/model_summary.md similarity index 94% rename from docs/user_manual/model/model_ summary.md rename to docs/user_manual/model/model_summary.md index 680847a..fe4fac0 100644 --- a/docs/user_manual/model/model_ summary.md +++ b/docs/user_manual/model/model_summary.md @@ -1,14 +1,14 @@ !!! Abstract "" 模型管理用于对接供应商的大语言模型，支持对接主流的大模型，包括本地私有大模型（Llama 3 / Qwen 2 等）、国内公共大模型（通义千问 / 智谱 AI / 百度千帆 / Kimi / DeepSeek 等）和国外公共大模型（OpenAI / Azure OpenAI / Gemini 等），支持的类型包括文本生成、向量模型、重排、语音识别、语音合成、图片理解、图片生成等。 - + 各供应商支持的类型详情见下表（按字母排序）： - | | 大语言模型 | 向量模型 | 重排模型 | 语音识别模型 | 语音合成模型 | 图片理解模型 | 图片生成模型 | + | | 大语言模型 | 向量模型 | 重排模型 | 语音识别模型 | 语音合成模型 | 图片理解模型 | 图片生成模型 | |-----------------|:--------:|:---------:|:----------:|:----------:|:----------:|:----------:|:----------:| - |阿里云百炼 | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | + |阿里云百炼 | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | |Amazon Bedrock | ✔ | ✔ | | | | | | |Azure OpenAI | ✔ | ✔ | | ✔ | ✔ | ✔ | ✔ | - |DeepSeek | ✔ | | | | | | | + |DeepSeek | ✔ | | | | | | | |Gemini | ✔ | ✔ | | ✔ | | ✔ | | |kimi-月之暗面 | ✔ | | | | | | | |Ollama | ✔ | ✔ | | | | ✔ | | @@ -25,4 +25,3 @@ !!! Abstract "" 登录 MaxKB 系统后，可以先对接模型，也可以在创建应用时再添加模型。添加模型时，可先点击【添加模型】，然后选择供应商，也可以左侧先选择供应商，然后点击【添加模型】。 - diff --git a/docs/user_manual/model/ollama_model.md b/docs/user_manual/model/ollama_model.md index 0b3decd..4e7da01 100644 --- a/docs/user_manual/model/ollama_model.md +++ b/docs/user_manual/model/ollama_model.md @@ -1,37 +1,37 @@ -## 1 添加模型 +## 1 Adding a Model -!!! Abstract "" - **说明：** Ollama允许用户在本地计算机上运行和部署大型语言模型。在添加 Ollama 模型之前需要先自行安装部署 Ollama 并运行模型，详情见：[Ollama 离线部署LLM模型](../../faq/Offline_install_OllamaModel.md)。 +!!! Abstract "" + **Note:** Ollama allows users to run and deploy large language models on their local computers. Before adding an Ollama model, you need to install and deploy Ollama and run the model. For more details, see: [Ollama Offline Deployment of LLM Models](../../faq/Offline_install_OllamaModel.md). -!!! Abstract "" - 模型供应商为`Ollama`，并在模型添加对话框中输入如下必要信息： +!!! Abstract "" + Select the model provider as `Ollama` and enter the following necessary information in the model addition dialog: - * 模型名称：MaxKB 中自定义的模型名称。 - * 权限：分为私有和公用两种权限，私有模型仅当前用户可用，公用模型即系统内所有用户均可使用，但其它用户不能编辑和删除。 - * 模型类型：大语言模型/向量模型/图片理解。 - * 基础模型：不同类型模型下的基础模型名称，下拉选项是常用的一些基础模型名称，支持自定义输入，但需要与 Ollama 支持的模型名称保持一致，否则无法通过校验。如果本地没有此模型，系统将自动下载。 + * Model Name: Custom model name in MaxKB. + * Permission: Divided into private and public permissions. Private models are only available to the current user, while public models can be used by all users within the system, but other users cannot edit or delete them. + * Model Type: Large language model/Vector model/Image understanding. + * Base Model: The name of the base model under different types, with a drop-down option for some commonly used base model names. It supports custom input, but must be consistent with the model names supported by Ollama; otherwise, it will not pass validation. If the model is not available locally, the system will automatically download it. -!!! Abstract "" - 大语言模型还需要输入 API 域名和API Key，向量模型需要输入API 域名。 +!!! Abstract "" + Large language models also require the input of an API domain and API Key, while vector models require the API domain. - * API 域名：为 Ollama 服务地址连接信息，例如：http://42.92.198.53:11434 。 - * API Key：若没有配置API Key，可以输入任意字符。 + * API Domain: Connection information for the Ollama service address, e.g., http://42.92.198.53:11434. + * API Key: If an API Key is not configured, you can enter any character. - 点击【添加】，校验通过则添加成功，便可以在应用的 AI 模型列表选择该模型。 + Click [Add], and if the validation passes, the model is successfully added. You will then be able to select this model in the AI model list of the application. -## 2 配置样例 +## 2 Configuration Example !!! Abstract "" - ollama-大语言模型配置样例图示： + Example of configuration for ollama - Large Language Model: -![ollama 大语言模型](../../img/model/ollama_model.png) +![ollama Large Language Model](../../img/model/ollama_model.png) !!! Abstract "" - ollama-向量模型配置样例图示： + Example of configuration for ollama - Vector Model: -![ollama 向量模型](../../img/model/ollama_embedding_model.png) +![ollama Vector Model](../../img/model/ollama_embedding_model.png) !!! Abstract "" - ollama-图片理解配置样例图示： + Example of configuration for ollama - Image Understanding: -![ollama 图片理解模型](../../img/model/ollama_vision.png) \ No newline at end of file +![ollama Image Understanding Model](../../img/model/ollama_vision.png) diff --git a/docs/user_manual/model/openai_model.md b/docs/user_manual/model/openai_model.md index 98bdb08..cb615a9 100644 --- a/docs/user_manual/model/openai_model.md +++ b/docs/user_manual/model/openai_model.md @@ -1,36 +1,33 @@ -## 1 添加模型 +## 1 Adding a Model !!! Abstract "" - 选择模型供应商为`OpenAI`，并在模型添加对话框中输入如下必要信息： + Select the model provider as `OpenAI` and enter the following necessary information in the model addition dialog: - * 模型名称： MaxKB 中自定义的模型名称。 - * 权限：分为私有和公用两种权限，私有模型仅当前用户可用，公用模型即系统内所有用户均可使用，但其它用户不能编辑和删除。 - * 模型类型： 大语言模型/向量模型/语音识别/语音合成。 - * 基础模型： 不同类型模型下的基础模型名称，下拉选项是常用的一些基础模型名称，支持自定义输入，但需要与 OpenAI 支持的模型名称保持一致，否则无法通过校验。 - * API 域名：国外 Open API的域名是`https://api.openai.com/v1`，国内代理的API域名格式一般是`反向代理地址/v1`。 - * API Key：访问 OpenAI 的 Key。 + * Model Name: Custom model name in MaxKB. + * Permission: Divided into private and public permissions. Private models are only available to the current user, while public models can be used by all users within the system, but other users cannot edit or delete them. + * Model Type: Large language model/Vector model/Speech recognition/Speech synthesis. + * Base Model: The name of the base model under different types, with a drop-down option for some commonly used base model names. It supports custom input but must be consistent with the model names supported by OpenAI; otherwise, it will not pass validation. + * API Domain: The domain name for the overseas Open API is `https://api.openai.com/v1`. The format for the API domain in China is typically `reverse proxy address/v1`. + * API Key: The Key to access OpenAI. -## 2 配置样例 +## 2 Configuration Example !!! Abstract "" - OpenAI-大语言模型配置样例图示： -![OpenAI 模型](../../img/model/openai_llm.png){ width="500px" } - + Example of OpenAI - Large Language Model configuration: +![OpenAI Model](../../img/model/openai_llm.png){ width="500px" } !!! Abstract "" - OpenAI-向量模型配置样例图示： -![OpenAI 模型](../../img/model/openai_embed.png){ width="500px" } - + Example of OpenAI - Vector Model configuration: +![OpenAI Model](../../img/model/openai_embed.png){ width="500px" } !!! Abstract "" - OpenAI-语音识别模型配置样例图示： -![OpenAI 模型](../../img/model/openai_asr.png){ width="500px" } - + Example of OpenAI - Speech Recognition Model configuration: +![OpenAI Model](../../img/model/openai_asr.png){ width="500px" } !!! Abstract "" - OpenAI-语音合成配置样例图示： -![OpenAI 模型](../../img/model/openai_tts.png){ width="500px" } + Example of OpenAI - Speech Synthesis configuration: +![OpenAI Model](../../img/model/openai_tts.png){ width="500px" } !!! Abstract "" - OpenAI-图片理解配置样例图示： -![OpenAI 模型](../../img/model/openai_vision.png){ width="500px" } \ No newline at end of file + Example of OpenAI - Image Understanding configuration: +![OpenAI Model](../../img/model/openai_vision.png){ width="500px" } diff --git a/docs/user_manual/model/qianfan_model.md b/docs/user_manual/model/qianfan_model.md deleted file mode 100644 index 16e4203..0000000 --- a/docs/user_manual/model/qianfan_model.md +++ /dev/null @@ -1,26 +0,0 @@ -## 1 添加模型 - -!!! Abstract "" - 添加千帆大模型之前，需要先在 [百度智能云千帆大模型平台](https://qianfan.cloud.baidu.com/) 中进行注册并登录。在控制台中的【模型服务-应用接入】中创建应用后，将自动生成 API Key和Secret Key 等信息。 -![百度应用](../../img/model/qianfan_app.png) - -!!! Abstract "" - 选择模型供应商为`千帆大模型`，并在模型添加对话框中输入如下必要信息： - - * 模型名称：MaxKB 中自定义的模型名称。 - * 权限：分为私有和公用两种权限，私有模型仅当前用户可用，公用模型即系统内所有用户均可使用，但其它用户不能编辑和删除。 - * 模型类型：大语言模型/向量模型。 - * 基础模型：百度千帆支持的 LLM 模型名称，选项中显示了百度千帆支持的部分常用大语言模型名称，支持手动输入，但需要与千帆大平台支持的模型名称保持一致，否则无法通过校验。 - * API Key 和 Secret Key：千帆大模型中应用的API Key 和 Secret Key。 - -## 2 配置样例 - -!!! Abstract "" - 千帆大模型-大语言模型配置样例图示： - -![千帆大模型](../../img/model/qianfan_model.png) - -!!! Abstract "" - 千帆大模型-向量模型配置样例图示： -![千帆大模型](../../img/model/qianfan_embedding.png) - diff --git a/docs/user_manual/model/spark_model.md b/docs/user_manual/model/spark_model.md deleted file mode 100644 index cb3c79a..0000000 --- a/docs/user_manual/model/spark_model.md +++ /dev/null @@ -1,48 +0,0 @@ -## 1 添加模型 - -!!! Abstract "" - 添加讯飞星火大模型之前，需要先在 [讯飞开放平台](https://www.xfyun.cn/) 中进行注册，并根据不同的模型类型创建不同应用，平台将自动生成APPID、API Secret、APIKey等信息。 - -![讯飞APP](../../img/model/xunfei_app.png) -![讯飞APP](../../img/model/xunfei_s2t_app.png) -![讯飞APP](../../img/model/xunfei_t2s_app.png) - - -!!! Abstract "" - 选择模型供应商为`讯飞星火`，并在模型添加对话框中输入如下必要信息： - - * 模型名称：MaxKB 中自定义的模型名称。 - * 权限：分为私有和公用两种权限，私有模型仅当前用户可用，公用模型即系统内所有用户均可使用，但其它用户不能编辑和删除。 - * 模型类型：大语言模型/向量模型/语音识别/语音合成。 - * 基础模型：对应接口文档中的 domain，下拉选项是讯飞星火不同模型类型下的常用基础模型名称，支持自定义输入。 - * API 域名：每个基础模型对应的 API 域名不同，请根据所选基础模型输入对应的 API 域名，详情请参考[讯飞开放平台文档中心](https://www.xfyun.cn/doc/)，如下图所示。 - * APP ID：讯飞开放平台中创建应用的标识。 - * API Key：讯飞开放平台应用访问对应的 API Key。 - * API Secret：讯飞开放平台应用访问对应的的 API Secret。 - - -![讯飞星火认知大模型 API](../../img/model/xunfei_api.png) -![讯飞语音识别大模型 API](../../img/model/xunfei_tts_wss.png) -![讯飞语音合成大模型 API](../../img/model/xunfei_iat_wss.png) - -## 2 配置样例 - -!!! Abstract "" - 讯飞星火-大语言模型配置样例图示： - -![讯飞模型](../../img/model/xunfei_llm.png){ width="500px" } - -!!! Abstract "" - 讯飞星火-向量模型配置样例图示： - -![讯飞模型](../../img/model/xunfei_embed.png){ width="500px" } - -!!! Abstract "" - 讯飞星火-语音识别模型配置样例图示： - -![讯飞模型](../../img/model/xunfei_iat.png){ width="500px" } - -!!! Abstract "" - 讯飞星火-语音合成模型配置样例图示： - -![讯飞模型](../../img/model/xunfei_tts.png){ width="500px" } \ No newline at end of file diff --git a/docs/user_manual/model/tongyi_model.md b/docs/user_manual/model/tongyi_model.md deleted file mode 100644 index e41c7b3..0000000 --- a/docs/user_manual/model/tongyi_model.md +++ /dev/null @@ -1,31 +0,0 @@ -## 1 添加模型 - -!!! Abstract "" - 添加通义千问模型之前，需要先在阿里云 DashScope 模型服务灵积的控制台中创建 API-Key。 - -![通义千问 APIkey](../../img/model/tongyi_apikey.png) - -!!! Abstract "" - 选择模型供应商为`通义千问`，并在模型添加对话框中输入如下必要信息： - - * 模型名称：MaxKB 中自定义的模型名称。 - * 权限：分为私有和公用两种权限，私有模型仅当前用户可用，公用模型即系统内所有用户均可使用，但其它用户不能编辑和删除。 - * 模型类型：大语言模型/图片理解/图片生成。 - * 基础模型：不同类型模型下支持的基础模型名称，下拉选项是常用的一些基础模型名称，支持自定义输入。 - * API Key：阿里云 DashScope灵积模型服务 API Key管理中创建和查看。 - -## 2 配置样例 - -!!! Abstract "" - 通义千问-大语言模型配置样例图示如下： -![通义千问 大语言模型](../../img/model/tongyi_llm.png){ width="500px" } - -!!! Abstract "" - 通义千问-图片理解模型配置样例图示如下： -![通义千问 图片模型](../../img/model/tongyi_vision.png){ width="500px" } - -!!! Abstract "" - 通义千问-图片生成模型默认图像尺寸为 1024 * 1024，图片数量 1 张，风格为 ，即由模型随机输出图像风格，配置样例图示如下： -![通义千问 图片模型](../../img/model/tongyi_vision_gen1.png){ width="500px" } - -![通义千问 图片模型](../../img/model/tongyi_vision_gen2.png){ width="500px" } diff --git a/docs/user_manual/model/vllm_model.md b/docs/user_manual/model/vllm_model.md index 6877d0c..1e0fca7 100644 --- a/docs/user_manual/model/vllm_model.md +++ b/docs/user_manual/model/vllm_model.md @@ -1,18 +1,18 @@ -## 1 添加模型 +## 1 Add Model !!! Abstract "" - 选择模型供应商为`vLLM`，并在模型添加对话框中输入如下必要信息： + Select the model provider as `vLLM`, and enter the following necessary information in the model addition dialog box: - * 模型名称：MaxKB 中自定义的模型名称。 - * 权限：分为私有和公用两种权限，私有模型仅当前用户可用，公用模型即系统内所有用户均可使用，但其它用户不能编辑和删除。 - * 模型类型：大语言模型。 - * 基础模型：不同类型模型下的基础模型名称，下拉选项是常用的一些基础模型名称，支持自定义输入。 - * API 域名：vLLM 服务地址， 如：http://192.168.20.242:8000/v1 。 - * API Key：若没有 API Key，输入任意字符即可。 + * Model Name: The custom model name in MaxKB. + * Permissions: Divided into private and public permissions. Private models are available only to the current user, while public models can be used by all users within the system, but other users cannot edit or delete them. + * Model Type: Large Language Model. + * Base Model: The base model name under different model types. The dropdown options include some commonly used base model names, and custom input is supported. + * API Domain: vLLM service address, e.g., http://192.168.20.242:8000/v1. + * API Key: If there is no API Key, you can enter any characters. -## 2 配置样例 +## 2 Configuration Example !!! Abstract "" - vLLM-大语言模型配置样例图示： + vLLM-Large Language Model configuration example illustration: -![vLLM LLM模型](../../img/model/vLLM_llm.png){ width="500px" } \ No newline at end of file +![vLLM LLM Model](../../img/model/vLLM_llm.png){ width="500px" } diff --git a/docs/user_manual/model/zhipu_model.md b/docs/user_manual/model/zhipu_model.md deleted file mode 100644 index c3d6384..0000000 --- a/docs/user_manual/model/zhipu_model.md +++ /dev/null @@ -1,34 +0,0 @@ -## 1 添加模型 - -!!! Abstract "" - 添加讯智谱AI模型之前，需要先在 [智谱AI开放平台](https://open.bigmodel.cn/) 中进行注册并创建 API Key。 - -![智普APIkey](../../img/model/zhipu_apikey.png) - -!!! Abstract "" - 选择模型供应商为`智谱AI`，并在模型添加对话框中输入如下必要信息： - - * 模型名称：MaxKB 中自定义的模型名称。 - * 权限：分为私有和公用两种权限，私有模型仅当前用户可用，公用模型即系统内所有用户均可使用，但其它用户不能编辑和删除。 - * 模型类型：大语言模型/图片理解/图片生成。 - * 基础模型：不同类型模型下的基础模型名称，下拉选项是常用的一些基础模型名称，支持自定义输入。 - * API Key：在智谱 AI 开放平台创建的 API Key。 - -## 2 配置样例 - -!!! Abstract "" - 智谱AI-大语言模型配置样例图示如下： - -![智普_大语言模型](../../img/model/zhipu_llm.png){ width="500px" } - -!!! Abstract "" - 智谱AI-图片理解模型配置样例图示如下： - -![智普_图片理解模型](../../img/model/zhipu_vision.png){ width="500px" } - -!!! Abstract "" - 智谱AI-图片生成模型配置样例图示如下： - -![智普_图片生成模型](../../img/model/zhipu_vision_gen1.png){ width="500px" } - -![智普_图片生成模型](../../img/model/zhipu_vision_gen2.png){ width="500px" } \ No newline at end of file diff --git a/mkdocs.yml b/mkdocs.yml index 99f2b30..8703f15 100644 --- a/mkdocs.yml +++ b/mkdocs.yml @@ -28,7 +28,7 @@ theme: text: Roboto code: Roboto Mono extra_css: - - stylesheets/extra.css + - stylesheets/extra.css markdown_extensions: - toc: permalink: true @@ -42,7 +42,7 @@ markdown_extensions: alternate_style: true ## grid - attr_list - - md_in_html + - md_in_html - tables - pymdownx.emoji: emoji_index: !!python/name:material.extensions.emoji.twemoji @@ -71,8 +71,52 @@ plugins: nav: - Introduction: index.md - Get Started: - - Quick Start: quick_start.md - - #- Tutorials: tutorials.md + - Quick Start: quick_start.md + - Installation Guide: + - Online Installation: installation/online_installtion.md + - 1Panel Installation: installation/1panel_installtion.md + - Command Line Tools Setup: installation/cli.md + - User Manual: + - Knowledge Base: + - Knowledge Base: user_manual/dataset/dataset.md + - Documents: user_manual/dataset/doclist.md + - Questions: user_manual/dataset/problem.md + - Hit Testing: user_manual/dataset/hit-testing.md + - Function Library: user_manual/fx/fx.md + - Applications: + - Application Creation: + - Simple Configuration Applications: user_manual/app/simple_app.md + - Advanced Orchestration Applications: user_manual/app/workflow_app.md + - Export and Import: user_manual/app/serialization.md + - Application Overview: user_manual/app/app-view.md + - Hit Testing: user_manual/app/hit-testing.md + - Conversation Logs: user_manual/app/log.md + - Model Management: + - Model Overview: user_manual/model/model_summary.md + - Model Integration: + - Integrate with Amazon Bedrock: user_manual/model/amazon_bedrock_model.md + - Integrate with Azure OpenAI: user_manual/model/azure_openai_model.md + - Integrate with Gemini: user_manual/model/gemini_model.md + - Integrate with OpenAI: user_manual/model/openai_model.md + - Integrate with Local Models: user_manual/model/local_model.md + - Integrate with Ollama: user_manual/model/ollama_model.md + - Integrate with vLLM: user_manual/model/vllm_model.md + - Integrate with Xorbits Inference: user_manual/model/X-Infer_model.md + - Model Parameter Configuration: user_manual/model/model_param.md + - Team Management: user_manual/team/team.md + - X-Pack Features: + - User Management: user_manual/X-Pack/user.md + - Login Authentication: user_manual/X-Pack/login_auth.md + - Application Display: user_manual/X-Pack/app_logo_settting.md + - Authentication: user_manual/X-Pack/app_auth.md + - System API: user_manual/X-Pack/system_API.md + - System Appearance: user_manual/X-Pack/theme_settings.md - Development: - - Building from Source: dev_manual/build_from_source.md + - Building from Source: dev_manual/build_from_source.md + - Chat via API KEY: dev_manual/APIKey_chat.md + - FAQ: + - Deploy LLM Model Offline with Ollama: faq/Offline_install_OllamaModel.md + - Run LLM Model with GPU using Ollama: faq/GPU_runOllama.md + - Integrate MaxKB Assistant into Halo: faq/maxkb_In_halo.md + - Difference Between MaxKB and Dify: faq/MaxKB_VS_Dify.md + - Changelog: changelog.md