[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-adieyal--sd-dynamic-prompts":3,"tool-adieyal--sd-dynamic-prompts":62},[4,18,26,35,44,53],{"id":5,"name":6,"github_repo":7,"description_zh":8,"stars":9,"difficulty_score":10,"last_commit_at":11,"category_tags":12,"status":17},4358,"openclaw","openclaw\u002Fopenclaw","OpenClaw 是一款专为个人打造的本地化 AI 助手，旨在让你在自己的设备上拥有完全可控的智能伙伴。它打破了传统 AI 助手局限于特定网页或应用的束缚，能够直接接入你日常使用的各类通讯渠道，包括微信、WhatsApp、Telegram、Discord、iMessage 等数十种平台。无论你在哪个聊天软件中发送消息，OpenClaw 都能即时响应，甚至支持在 macOS、iOS 和 Android 设备上进行语音交互，并提供实时的画布渲染功能供你操控。\n\n这款工具主要解决了用户对数据隐私、响应速度以及“始终在线”体验的需求。通过将 AI 部署在本地，用户无需依赖云端服务即可享受快速、私密的智能辅助，真正实现了“你的数据，你做主”。其独特的技术亮点在于强大的网关架构，将控制平面与核心助手分离，确保跨平台通信的流畅性与扩展性。\n\nOpenClaw 非常适合希望构建个性化工作流的技术爱好者、开发者，以及注重隐私保护且不愿被单一生态绑定的普通用户。只要具备基础的终端操作能力（支持 macOS、Linux 及 Windows WSL2），即可通过简单的命令行引导完成部署。如果你渴望拥有一个懂你",349277,3,"2026-04-06T06:32:30",[13,14,15,16],"Agent","开发框架","图像","数据工具","ready",{"id":19,"name":20,"github_repo":21,"description_zh":22,"stars":23,"difficulty_score":10,"last_commit_at":24,"category_tags":25,"status":17},3808,"stable-diffusion-webui","AUTOMATIC1111\u002Fstable-diffusion-webui","stable-diffusion-webui 是一个基于 Gradio 构建的网页版操作界面，旨在让用户能够轻松地在本地运行和使用强大的 Stable Diffusion 图像生成模型。它解决了原始模型依赖命令行、操作门槛高且功能分散的痛点，将复杂的 AI 绘图流程整合进一个直观易用的图形化平台。\n\n无论是希望快速上手的普通创作者、需要精细控制画面细节的设计师，还是想要深入探索模型潜力的开发者与研究人员，都能从中获益。其核心亮点在于极高的功能丰富度：不仅支持文生图、图生图、局部重绘（Inpainting）和外绘（Outpainting）等基础模式，还独创了注意力机制调整、提示词矩阵、负向提示词以及“高清修复”等高级功能。此外，它内置了 GFPGAN 和 CodeFormer 等人脸修复工具，支持多种神经网络放大算法，并允许用户通过插件系统无限扩展能力。即使是显存有限的设备，stable-diffusion-webui 也提供了相应的优化选项，让高质量的 AI 艺术创作变得触手可及。",162132,"2026-04-05T11:01:52",[14,15,13],{"id":27,"name":28,"github_repo":29,"description_zh":30,"stars":31,"difficulty_score":32,"last_commit_at":33,"category_tags":34,"status":17},2271,"ComfyUI","Comfy-Org\u002FComfyUI","ComfyUI 是一款功能强大且高度模块化的视觉 AI 引擎，专为设计和执行复杂的 Stable Diffusion 图像生成流程而打造。它摒弃了传统的代码编写模式，采用直观的节点式流程图界面，让用户通过连接不同的功能模块即可构建个性化的生成管线。\n\n这一设计巧妙解决了高级 AI 绘图工作流配置复杂、灵活性不足的痛点。用户无需具备编程背景，也能自由组合模型、调整参数并实时预览效果，轻松实现从基础文生图到多步骤高清修复等各类复杂任务。ComfyUI 拥有极佳的兼容性，不仅支持 Windows、macOS 和 Linux 全平台，还广泛适配 NVIDIA、AMD、Intel 及苹果 Silicon 等多种硬件架构，并率先支持 SDXL、Flux、SD3 等前沿模型。\n\n无论是希望深入探索算法潜力的研究人员和开发者，还是追求极致创作自由度的设计师与资深 AI 绘画爱好者，ComfyUI 都能提供强大的支持。其独特的模块化架构允许社区不断扩展新功能，使其成为当前最灵活、生态最丰富的开源扩散模型工具之一，帮助用户将创意高效转化为现实。",108111,2,"2026-04-08T11:23:26",[14,15,13],{"id":36,"name":37,"github_repo":38,"description_zh":39,"stars":40,"difficulty_score":32,"last_commit_at":41,"category_tags":42,"status":17},4721,"markitdown","microsoft\u002Fmarkitdown","MarkItDown 是一款由微软 AutoGen 团队打造的轻量级 Python 工具，专为将各类文件高效转换为 Markdown 格式而设计。它支持 PDF、Word、Excel、PPT、图片（含 OCR）、音频（含语音转录）、HTML 乃至 YouTube 链接等多种格式的解析，能够精准提取文档中的标题、列表、表格和链接等关键结构信息。\n\n在人工智能应用日益普及的今天，大语言模型（LLM）虽擅长处理文本，却难以直接读取复杂的二进制办公文档。MarkItDown 恰好解决了这一痛点，它将非结构化或半结构化的文件转化为模型“原生理解”且 Token 效率极高的 Markdown 格式，成为连接本地文件与 AI 分析 pipeline 的理想桥梁。此外，它还提供了 MCP（模型上下文协议）服务器，可无缝集成到 Claude Desktop 等 LLM 应用中。\n\n这款工具特别适合开发者、数据科学家及 AI 研究人员使用，尤其是那些需要构建文档检索增强生成（RAG）系统、进行批量文本分析或希望让 AI 助手直接“阅读”本地文件的用户。虽然生成的内容也具备一定可读性，但其核心优势在于为机器",93400,"2026-04-06T19:52:38",[43,14],"插件",{"id":45,"name":46,"github_repo":47,"description_zh":48,"stars":49,"difficulty_score":10,"last_commit_at":50,"category_tags":51,"status":17},4487,"LLMs-from-scratch","rasbt\u002FLLMs-from-scratch","LLMs-from-scratch 是一个基于 PyTorch 的开源教育项目，旨在引导用户从零开始一步步构建一个类似 ChatGPT 的大型语言模型（LLM）。它不仅是同名技术著作的官方代码库，更提供了一套完整的实践方案，涵盖模型开发、预训练及微调的全过程。\n\n该项目主要解决了大模型领域“黑盒化”的学习痛点。许多开发者虽能调用现成模型，却难以深入理解其内部架构与训练机制。通过亲手编写每一行核心代码，用户能够透彻掌握 Transformer 架构、注意力机制等关键原理，从而真正理解大模型是如何“思考”的。此外，项目还包含了加载大型预训练权重进行微调的代码，帮助用户将理论知识延伸至实际应用。\n\nLLMs-from-scratch 特别适合希望深入底层原理的 AI 开发者、研究人员以及计算机专业的学生。对于不满足于仅使用 API，而是渴望探究模型构建细节的技术人员而言，这是极佳的学习资源。其独特的技术亮点在于“循序渐进”的教学设计：将复杂的系统工程拆解为清晰的步骤，配合详细的图表与示例，让构建一个虽小但功能完备的大模型变得触手可及。无论你是想夯实理论基础，还是为未来研发更大规模的模型做准备",90106,"2026-04-06T11:19:32",[52,15,13,14],"语言模型",{"id":54,"name":55,"github_repo":56,"description_zh":57,"stars":58,"difficulty_score":10,"last_commit_at":59,"category_tags":60,"status":17},4292,"Deep-Live-Cam","hacksider\u002FDeep-Live-Cam","Deep-Live-Cam 是一款专注于实时换脸与视频生成的开源工具，用户仅需一张静态照片，即可通过“一键操作”实现摄像头画面的即时变脸或制作深度伪造视频。它有效解决了传统换脸技术流程繁琐、对硬件配置要求极高以及难以实时预览的痛点，让高质量的数字内容创作变得触手可及。\n\n这款工具不仅适合开发者和技术研究人员探索算法边界，更因其极简的操作逻辑（仅需三步：选脸、选摄像头、启动），广泛适用于普通用户、内容创作者、设计师及直播主播。无论是为了动画角色定制、服装展示模特替换，还是制作趣味短视频和直播互动，Deep-Live-Cam 都能提供流畅的支持。\n\n其核心技术亮点在于强大的实时处理能力，支持口型遮罩（Mouth Mask）以保留使用者原始的嘴部动作，确保表情自然精准；同时具备“人脸映射”功能，可同时对画面中的多个主体应用不同面孔。此外，项目内置了严格的内容安全过滤机制，自动拦截涉及裸露、暴力等不当素材，并倡导用户在获得授权及明确标注的前提下合规使用，体现了技术发展与伦理责任的平衡。",88924,"2026-04-06T03:28:53",[14,15,13,61],"视频",{"id":63,"github_repo":64,"name":65,"description_en":66,"description_zh":67,"ai_summary_zh":67,"readme_en":68,"readme_zh":69,"quickstart_zh":70,"use_case_zh":71,"hero_image_url":72,"owner_login":73,"owner_name":74,"owner_avatar_url":75,"owner_bio":76,"owner_company":76,"owner_location":76,"owner_email":76,"owner_twitter":76,"owner_website":76,"owner_url":77,"languages":78,"stars":95,"forks":96,"last_commit_at":97,"license":98,"difficulty_score":32,"env_os":99,"env_gpu":99,"env_ram":99,"env_deps":100,"category_tags":104,"github_topics":76,"view_count":32,"oss_zip_url":76,"oss_zip_packed_at":76,"status":17,"created_at":105,"updated_at":106,"faqs":107,"releases":137},5585,"adieyal\u002Fsd-dynamic-prompts","sd-dynamic-prompts","A custom script for AUTOMATIC1111\u002Fstable-diffusion-webui to implement a tiny template language for random prompt generation","sd-dynamic-prompts 是一款专为 AUTOMATIC1111 Stable Diffusion WebUI 设计的扩展插件，旨在通过引入灵活的模板语言，实现提示词（Prompt）的随机化与组合式生成。它有效解决了用户在创作中难以快速探索海量风格组合、手动编写大量变体提示词效率低下的痛点，让单次输入即可衍生出无数种创意可能。\n\n无论是希望激发灵感的设计师、需要批量测试模型效果的研究人员，还是想要轻松尝试不同艺术风格的普通爱好者，都能从中受益。其核心亮点在于简洁直观的语法：用户只需使用花括号 `{选项 A|选项 B}` 即可定义随机选择，或利用 `__文件名__` 调用外部通配符库（如艺术家名单、季节词汇等），甚至支持从深层目录结构中递归读取资源。此外，它还具备“组合生成模式”，能系统性地将所有选项排列组合并批量出图，极大提升了实验效率。配合内置的通配符管理器和丰富的在线资源社区，sd-dynamic-prompts 让提示词工程变得更加简单有趣，是挖掘 Stable Diffusion 潜能的得力助手。","# Stable Diffusion Dynamic Prompts extension\nA custom extension for [AUTOMATIC1111\u002Fstable-diffusion-webui](https:\u002F\u002Fgithub.com\u002FAUTOMATIC1111\u002Fstable-diffusion-webui) that implements an expressive template language for random or combinatorial prompt generation along with features to support deep wildcard directory structures.\n\nLooking for ComfyUI nodes? Find them [here](https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fcomfyui-dynamicprompts).\n\n\u003Cp align=\"center\">\n    \u003Ca href=\"docs\u002FSYNTAX.md\">\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_5c08bc0775c8.png\" valign=\"middle\" style=\"height:60px;margin-right:10px\"\u002F>\u003C\u002Fa>\n    \u003Ca href=\"docs\u002Ftutorial.md\">\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_c9daf1469ba1.png\" valign=\"middle\" style=\"height:60px;margin-right:10px\"\u002F>\u003C\u002Fa>\n    \u003Ca href=\"docs\u002FCHANGELOG.md\">\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_a1a23345b43f.png\" valign=\"middle\" style=\"height:60px\"\u002F>\u003C\u002Fa>\n\u003C\u002Fp>\n\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_fed9dfd22d46.png\"\u002F>\n\n## Table of Contents\n\n   * [Basic Usage](#basic-usage)\n   * [Online resources](#online-resources)\n   * [Installation](#installation)\n   * [Configuration](#configuration)\n   * [Troubleshooting](#troubleshooting)\n   * [Compatible Scripts](#compatible-scripts)\n   * [Template syntax](#template-syntax)\n      * [Fuzzy Glob\u002Frecursive wildcard file\u002Fdirectory matching](#fuzzy-globrecursive-wildcard-filedirectory-matching)\n   * [Combinatorial Generation](#combinatorial-generation)\n      * [Combinatorial Batches](#combinatorial-batches)\n      * [Increasing the maximum number of generations](#increasing-the-maximum-number-of-generations)\n   * [Fixed seed](#fixed-seed)\n   * [Magic Prompt](#magic-prompt)\n      * [Other models](#other-models)\n   * [I'm feeling lucky](#im-feeling-lucky)\n   * [Attention grabber](#attention-grabber)\n   * [Write prompts to file](#write-prompts-to-file)\n   * [Jinja2 templates](#jinja2-templates)\n   * [WILDCARD_DIR](#wildcard_dir)\n   * [Collections](#collections)\n   * [Dynamic Prompts and Random Seeds](#dynamic-prompts-and-random-seeds)\n      * [Without Dynamic Prompts Enabled](#without-dynamic-prompts-enabled)\n      * [With Dynamic Prompts Enabled in Random\u002FStandard Mode](#with-dynamic-prompts-enabled-in-randomstandard-mode)\n      * [Variation Seeds with Dynamic Prompts](#variation-seeds-with-dynamic-prompts)\n      * [Combinatorial Mode with Variation Strength > 0](#combinatorial-mode-with-variation-strength--0)\n\n\n## Basic Usage\nUsing this script, the prompt:\n\n\tA {house|apartment|lodge|cottage} in {summer|winter|autumn|spring} by {2$$artist1|artist2|artist3}\n\nWill produce any of the following prompts:\n\n- A **house** in **summer** by **artist1**, **artist2**\n- A **lodge** in **autumn** by **artist3**, **artist1**\n- A **cottage** in **winter** by **artist2**, **artist3**\n- ...\n\nThis is especially useful if you are searching for interesting combinations of artists and styles.\n\nYou can also pick a random string from a file. Assuming you have the file seasons.txt in WILDCARD_DIR (see below), then:\n\n    __seasons__ is coming\n\nMight generate the following:\n\n- Winter is coming\n- Spring is coming\n- ...\n\nYou can also use the same wildcard twice\n\n    I love __seasons__ better than __seasons__\n\n- I love Winter better than Summer\n- I love Spring better than Spring\n\nMore complete documentation can be found [here](docs\u002FSYNTAX.md).\u003Cbr\u002F>\nPrefer a tutorial? \u003Ca href=\"docs\u002Ftutorial.md\">Click here\u003C\u002Fa>\u003Cbr\u002F>\nNeed a wildcard library? We've got you [covered](#collections).\n\n## Online resources\nYou can find a list of tutorials and wildcard packs [here](docs\u002Fresources.md)\n\n## Installation\n\nThe extension can be installed directly from within the **Extensions** tab within the Webui\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_4cb7c603b44c.png\"\u002F>\n\nYou can also install it manually by running the following command from within the webui directory:\n\n```shell\ngit clone https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\u002F extensions\u002Fsd-dynamic-prompts\n```\n\nYou can create [wildcard files](#template-syntax) in `extensions\u002Fsd-dynamic-prompts\u002Fwildcards` or you can leverage the [pre-installed collections](#collections)\nby using the \"Wildcards Manager\" tab in the Web UI.\n\n## Configuration\nYou can find various settings to change Dynamic Prompt's behaviour in the Settings tab in the Dynamic Prompts section.\n\n### Changing syntax\nIn case of a syntax clash with another extension, Dynamic Prompts allows you to change the definition of variant start and variant end. By default these are set to `{` and `}` respectively. , e.g. `{red|green|blue}`. In the settings tab, you can change these two any string, e.g. `\u003Cred|green|blue>` or even `::red|green|blue::`.\n\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_8edfeea57f1c.png\">\n\nBy default, wildcards start with `__`(double underscore) and end with `__`. You can change this in the settings tab under wildcard wrap.\n\u003Cbr\u002F>\u003Cbr\u002F>\n\n### Wildcard settings\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_6d498d91d25b.png\">\n\nDynamic Prompts automatically de-duplicates and sorts wildcard files before using them. If you would prefer to disable this functionality, you can uncheck the checkboxes in the settings tab.\n\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_4d7882be46ab.png\">\n\nChecking the \"shuffle wildcards\" checkbox will randomize the order of the wildcards, ensuring that running the combinatorial model will produce different images on different runs.\n\n### Save template to metadata\n\nThe default behavior is to resolve all the wildcards to a usable prompt, then that prompt is stored in the PNG info or txt file (e.g. `I love Winter better than Summer`).  If you would also\nlike to recall the original template with the wildcards you can turn this option on and you'll save:\n\n```\nI love Winter better than Summer\nTemplate: I love __seasons__ better than __seasons__\n```\n\n> Note: this additional \"Template\" section is not displayed in the generate page but will be available in PNG Info (or image browser, if you have that extension installed).\n\n## Troubleshooting\nIf you encounter an issue with Dynamic Prompts, follow these steps to resolve the problem:\n\n1. Check that you have installed the latest version of both the Dynamic Prompts extension and library. To check the installed versions, open the **Need Help? accordion** in the Dynamic Prompts section of txt2image. You can find the latest version number of the extension [here](https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\u002Fblob\u002Fmain\u002Fdocs\u002FCHANGELOG.md) and the library [here](https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fdynamicprompts\u002Fblob\u002Fmain\u002FCHANGELOG.md?plain=1).\n\n2. If the versions do not match, update the extension in the extensions tab and restart the webui. The extension should automatically update the library.\n\n3. If the above step does not work, you might need to manually update the library using the following command:\n\n```shell\npython -m pip install -U dynamicprompts[attentiongrabber,magicprompt]\n```\n\n4. Restart the webui and check. If the webui uses a different python binary, find the correct path to the python binary and run:\n\n```shell\n\u002Fpath\u002Fto\u002Fpython\u002Fbinary\u002Fpython -m pip install -U dynamicprompts[attentiongrabber,magicprompt]\n```\n\n5. If the Wildcard UI does not show, it could be due to an outdated library version. Check for errors in the terminal and update the library as described in step 3.\n\n6. If you get an error message saying \"No values found for wildcard some\u002Fwildcard\", ensure that the file wildcard.txt is in extensions\u002Fsd-dynamic-prompts\u002Fwildcards\u002Fsome\u002F. The full path is required, as relative paths are not currently supported.\n\n7. If the issue persists, search for solutions in the [issues section](https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\u002Fissues?q=is%3Aissue) on GitHub and the [discussion forum](https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\u002Fdiscussions). If you cannot find a solution, create a new issue and give it a descriptive name, such as \"Wildcard values are being ignored in prompt templates\". Provide the necessary context, including the versions of the Dynamic Prompts extension and library, and mention the operating system or colab being used. If there is an error in the terminal, copy and paste the entire text or take a screenshot.\n\n8. Finally, it is essential to test and apply any fixes we release. Your feedback is valuable, as an issue that works in our environment may not work in yours.\n\n## Compatible Scripts\nDynamic Prompts works particularly well with X\u002FY Plot - setting Dynamic Prompts to \u003Ca href=\"#combinatorial-generation\">combinatorial mode\u003C\u002Fa> while using X\u002FY Plot, lets you exhaustively test prompt and paramter variations simultaneously.\n\n\n## Template syntax\nDocumentation can be found [here](docs\u002FSYNTAX.md)\n\n### Fuzzy Glob\u002Frecursive wildcard file\u002Fdirectory matching\nIn addition to standard wildcard tokens such as `__times__` -> `times.txt`, you can also use globbing to match against multiple files at once.\nFor instance:\n\n> `__colors*__` will match any of the following:\n> - WILDCARD_DIR\u002Fcolors.txt\n> - WILDCARD_DIR\u002Fcolors1.txt\n> - WILDCARD_DIR\u002Fnested\u002Ffolder\u002Fcolors1.txt\n>\n> `__light\u002F**\u002F*__` will match:\n> - WILDCARD_DIR\u002Fnested\u002Ffolder\u002Flight\u002Fa.txt\n> - WILDCARD_DIR\u002Fnested\u002Ffolder\u002Flight\u002Fb.txt\n>\n> but won't match\n> - WILDCARD_DIR\u002Fnested\u002Ffolder\u002Fdark\u002Fa.txt\n> - WILDCARD_DIR\u002Fa.txt\n\nYou can also used character ranges `[0-9]` and `[a-z]` and single wildcard characters `?`. For more examples see [this article](http:\u002F\u002Fpymotw.com\u002F2\u002Fglob\u002F).\n\n`WILDCARD_DIR` defaults to `extensions\u002Fsd-dynamic-prompts\u002Fwildcards`.\n\n## Combinatorial Generation\nInstead of generating random prompts from a template, combinatorial generation produced every possible prompt from the given string. For example:\n`I {love|hate} {New York|Chicago} in {June|July|August}`\n\nwill produce:\n- I love New York in June\n- I love New York in July\n- I love New York in August\n- I love Chicago in June\n- I love Chicago in July\n- I love Chicago in August\n- I hate New York in June\n- I hate New York in July\n- I hate New York in August\n- I hate Chicago in June\n- I hate Chicago in July\n- I hate Chicago in August\n\nIf a `__wildcard__` is provided, then a new prompt will be produced for every value in the wildcard file. For example:\n`My favourite season is __seasons__`\n\nwill produce:\n- My favourite season is Summer\n- My favourite season is August\n- My favourite season is Winter\n- My favourite season is Sprint\n\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_0e402ecc982d.png\"\u002F>\n\nYou also arbitrarily nest combinations inside wildcards and wildcards in combinations.\n\nCombinatorial generation can be useful if you want to create an image for every artist in a file. It can be enabled by checking the __Combinatorial generation__ checkbox in the ui. In order to prevent accidentially producing thousands of images, you can limit the total number of prompts generated using the **Max Generations** slider. A value of 0 (the default) will not set any limit.\n\n### Combinatorial Batches\nThe combinatorial batches slider lets you repeat the same set of prompts a number of times with different seeds. The default number of batches is 1.\n\n### Increasing the maximum number of generations\nBy default, the __Batch count__ silder of  automatic1111 has a maximum value of 100. This can limit the maximum number of generations when using combinatorial generation. You can change the maximum value of this slider by editing ui-config.json and change:\n\n\ttxt2img\u002FBatch count\u002Fmaximum\": 100\n\nto something larger like:\n\n\ttxt2img\u002FBatch count\u002Fmaximum\": 1000\n\n## Fixed seed\nSelect this if you want to use the same seed for every generated image. If there are no wildcards then all the images will be identical. It is useful if you want to test the effect of a particular modifier. For example:\n\n\tA beautiful day at the beach __medium\u002Fphotography\u002Ffilmtypes__\n\nThat way you can isolate the effect of each film type on a particular scene. Here are some of the results:\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_bd69d181a1bd.jpg\"\u002F>\n\n## Magic Prompt\nUse one of a number of prompt generation models to spice up your prompt.\n\nUsing [Gustavosta](https:\u002F\u002Fhuggingface.co\u002FGustavosta\u002FMagicPrompt-Stable-Diffusion)'s MagicPrompt model, Trained on 80,000 prompts from [Lexica.art](lexica.art), it can help give you interesting new prompts on a given subject. Here are some automatically generated variations for \"dogs playing football\":\n\n> dogs playing football, in the streets of a japanese town at night, with people watching in wonder, in the style of studio ghibli and makoto shinkai, highly detailed digital art, trending on artstation\n\n> dogs playing football, in the background is a nuclear explosion. photorealism. hq. hyper. realistic. 4 k. award winning.\n\n> dogs playing football, in the background is a nuclear explosion. photorealistic. realism. 4 k wideshot. cinematic. unreal engine. artgerm. marc simonetti. jc leyendecker\n\nThis is compatible with the wildcard syntax described above.\n\n### Other models\n* [daspartho\u002Fprompt-extend (~500mb)](https:\u002F\u002Fhuggingface.co\u002Fdaspartho\u002Fprompt-extend)\n* [succinctly\u002Ftext2image-prompt-generator (~600mb)](https:\u002F\u002Fhuggingface.co\u002Fsuccinctly\u002Ftext2image-prompt-generator) - Trained on Midjourney prompts\n* [microsoft\u002FPromptist (~500mb)](https:\u002F\u002Fhuggingface.co\u002Fmicrosoft\u002FPromptist) - Read the paper [here](https:\u002F\u002Farxiv.org\u002Fabs\u002F2212.09611)\n* [AUTOMATIC\u002Fpromptgen-lexart (~300mb)](https:\u002F\u002Fhuggingface.co\u002FAUTOMATIC\u002Fpromptgen-lexart) - Finetuned using 134,819 prompts from lexica.art\n* [AUTOMATIC\u002Fpromptgen-majinai-safe (~300mb)](https:\u002F\u002Fhuggingface.co\u002FAUTOMATIC\u002Fpromptgen-majinai-safe) - 1,654 prompts from majinai.art\n* [AUTOMATIC\u002Fpromptgen-majinai-unsafe (~300mb)](https:\u002F\u002Fhuggingface.co\u002FAUTOMATIC\u002Fpromptgen-majinai-unsafe) - 825 prompts from majinai.art (NSFW)\n* [Gustavosta\u002FMagicPrompt-Dalle](https:\u002F\u002Fhuggingface.co\u002FGustavosta\u002FMagicPrompt-Dalle)\n* [kmewhort\u002Fstable-diffusion-prompt-bolster (~500mb)](https:\u002F\u002Fhuggingface.co\u002Fkmewhort\u002Fstable-diffusion-prompt-bolster),\n* [Ar4ikov\u002Fgpt2-650k-stable-diffusion-prompt-generator (~500mb)](Ar4ikov\u002Fgpt2-650k-stable-diffusion-prompt-generator),\n* [Ar4ikov\u002Fgpt2-medium-650k-stable-diffusion-prompt-generator (~1.4gb)](https:\u002F\u002Fhuggingface.co\u002FAr4ikov\u002Fgpt2-medium-650k-stable-diffusion-prompt-generator),\n* [crumb\u002Fbloom-560m-RLHF-SD2-prompter-aesthetic (~1.1gb)](https:\u002F\u002Fhuggingface.co\u002Fcrumb\u002Fbloom-560m-RLHF-SD2-prompter-aesthetic),\n* [Meli\u002FGPT2-Prompt (~500mb)](https:\u002F\u002Fhuggingface.co\u002FMeli\u002FGPT2-Prompt),\n* [DrishtiSharma\u002FStableDiffusion-Prompt-Generator-GPT-Neo-125M (~550mb)](https:\u002F\u002Fhuggingface.co\u002FDrishtiSharma\u002FStableDiffusion-Prompt-Generator-GPT-Neo-125M)\n\nThe first time you use a model, it is downloaded. It is approximately 500mb and so will take some time depending on how fast your connection is. It will also take a few seconds on first activation as the model is loaded into memory. Note, if you're low in VRAM, you might get a Cuda error. My GPU uses less than 8GB but YMMV.\n\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_28f803324a6b.png\"\u002F>\n\n\nYou can control the maximum prompt length with the __Max magic prompt length__ slider. __Magic prompt creativity__ can adjust the generated prompt but you will need to experiment with this setting.\n\nUse the __Magic prompt blocklist regex__ to filter out keywords. For example, if you want to avoid prompts containing Greg Rutkowski, add his name to this field.\n\nIf you are generating many prompts using Magic Prompt, then increasing the __Magic Prompt batch size__ can improve significantly improve prompt generation speed. This may only be noticeable if you are not generating images as well since image generation is much slower than prompt generation.\n\n\n## I'm feeling lucky\nUse the [lexica.art](https:\u002F\u002Flexica.art) API to create random prompts. Useful if you're looking for inspiration, or are simply too lazy to think of your own prompts. When this option is selected, the prompt in the main prompt box is used as a search string. For example, prompt \"Mech warrior\" might return:\n\n* A large robot stone statue in the middle of a forest by Greg Rutkowski, Sung Choi, Mitchell Mohrhauser, Maciej Kuciara, Johnson Ting, Maxim Verehin, Peter Konig, final fantasy , 8k photorealistic, cinematic lighting, HD, high details, atmospheric,\n* a beautiful portrait painting of a ( ( ( cyberpunk ) ) ) armor by simon stalenhag and pascal blanche and alphonse mucha and nekro. in style of digital art. colorful comic, film noirs, symmetry, brush stroke, vibrating colors, hyper detailed. octane render. trending on artstation\n* symmetry!! portrait of a robot astronaut, floral! horizon zero dawn machine, intricate, elegant, highly detailed, digital painting, artstation, concept art, smooth, sharp focus, illustration, art by artgerm and greg rutkowski and alphonse mucha, 8 k\n\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_761cfcf6fc59.png\">\n\nLeaving the prompt box blank returns a list of completely randomly chosen prompts.\n\n## Attention grabber\nThis option randomly selects a keyword in your prompt and adds a random amount of emphasis. Below is an example of how this affects the prompt:\n\n\ta portrait an anthropomorphic panda mage casting a spell, wearing mage robes, landscape in background, cute, dnd character art portrait, by jason felix and peter mohrbacher, cinematic lighting\n\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_ee751c9a1162.png\">\n\nTick the __Fixed seed__ checkbox under __Advanced options__ to see how emphasis changes your image without changing seed.\n\n\n## Write prompts to file\nCheck the write prompts to file checkbox in order to create a file with all generated prompts. The generated file is a slugified version of the prompt and can be found in the same directory as the generated images, e.g. outputs\u002Ftxt2img-images\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_8dfe6d623df5.png\"\u002F>\n\n## Jinja2 templates\n[Jinja2 templates](https:\u002F\u002Fjinja.palletsprojects.com\u002Fen\u002F3.1.x\u002Ftemplates\u002F) is an experimental feature that enables you to define prompts imperatively. This is an advanced feature and is only recommended for users who are comfortable writing scripts.\n\nTo enable, open the advanced accordion and select __Enable Jinja2 templates__.\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_37b92400c8a9.png\">\n\nYou can read about them in more detail \u003Ca href=\"jinja2.md\">here\u003C\u002Fa>\n\n## WILDCARD_DIR\nThe extension looks for wildcard files in WILDCARD_DIR. The default location is `\u002Fpath\u002Fto\u002Fstable-diffusion-webui\u002Fextensions\u002Fsd-dynamic-prompts\u002Fwildcards`. It can also be manually defined in the main webui config.json under wildcard_dir. When in doubt, the help text for the extension in the webui lists the full path to WILDCARD_DIR\n\n## Collections\nThe collections directory contains modifier libraries that you can use as is or to bootstrap your own. To get started, either use the Wildcard Manager tab to copy a one or more collections to your wildcards folder, or you can manually copy the files across. Three collections are bundled with the dynamic prompts extension.\n\n- [jumbo](https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\u002Ftree\u002Fmain\u002Fcollections\u002Fjumbo)\n  A very large collection of wildcards across many categories including aesthetics, appearance, artists, medium, style, and time. It is a work in progress, but aims to provide good coverage of various modifier categories.\n- [parrotzone](https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\u002Ftree\u002Fmain\u002Fcollections\u002Fparrotzone)\n  A far smaller and more manageable collection sourced from https:\u002F\u002Fproximacentaurib.notion.site\u002Fe28a4f8d97724f14a784a538b8589e7d?v=42948fd8f45c4d47a0edfc4b78937474.\n- [devilkkw](https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\u002Ftree\u002Fmain\u002Fcollections\u002Fdevilkkw)\n  Devilkkw focuses more on character building, clothes, gestures, food, etc.\n\nIf you're using a Unix\u002FLinux O\u002FS, you can easily create a symlink to the relevant collection rather than copying it across if you don't plan to alter it. E.g.\n\n```shell\nln -sr collections\u002Fparrotzone wildcards\u002F\n```\n\nYou can also download additional extensions by running `python _tools\u002Fdownload_collections.py` from within the extension's root directory, i.e. `extensions\u002Fsd-dynamic-prompts\u002F`\n\n## Dynamic Prompts and Random Seeds\nRandom seeds play an important role in controlling the randomness of the generated outputs. Let's discuss how Dynamic Prompts works with random seeds in different scenarios.\n\n### Without Dynamic Prompts Enabled\n\n1. If the seed is set to -1: A random seed is picked. This seed is used to generate the first image, then the next image is created using seed + 1, and this pattern continues for subsequent images.\n2. If the seed is set to a specific number greater than -1: The process is similar to the one above, but starts with the user-specified seed.\n3. If the variation seed is defined, but variation strength is zero: The process remains the same as in the previous two points.\n4. If the variation seed is set to a number greater than 0: Every image is generated using the same initial seed (randomly selected or set by the user). The variation seed is either random (if set to -1) or the value chosen by the user. The first image is generated with the variation seed, the next with variation seed + 1, and so on.\n\n### Using With Dynamic Prompts Enabled in Random\u002FStandard Mode:\n\n1. If the seed is set to -1: The process is similar to the first point in the previous section. However, the prompt is also selected using the same seed (if the random prompt generator is used).\n2. If the seed is set to a number greater than -1: The process is similar to the second point in the previous section. However, the difference is that a random prompt is also generated using the chosen seed (if the prompt generator is used).\n3. If the fixed seed checkbox is checked: The same seed is used for all images and prompts. This means the same image is generated repeatedly (this is useful for combinatorial generation).\n4. If both the fixed seed and unlink seed from prompt checkboxes are checked: A random seed is used for the prompt, but the same seed is used for all images. This setting can be useful if you want to see how different prompts affect the generation of the same image.\n\n### Variation Seeds with Dynamic Prompts\n\n1. Variation strength set to 0: Variations are ignored.\n2. Variation set to a number > 0: A variation seed is assigned to every image, incrementing by one each time. However, only 1 prompt is generated since you are looking for variations of the same image.\n\n### Combinatorial Mode with Variation Strength > 0\nIn this case, it only generates the first image for you, which is probably not what you want. To get the desired results, you might need to adjust the settings or use a different mode.\n","# Stable Diffusion 动态提示词扩展\n这是一个为 [AUTOMATIC1111\u002Fstable-diffusion-webui](https:\u002F\u002Fgithub.com\u002FAUTOMATIC1111\u002Fstable-diffusion-webui) 开发的自定义扩展，它实现了一种功能强大的模板语言，用于随机或组合式提示词生成，并支持深度通配符目录结构。\n\n正在寻找 ComfyUI 节点？请在此处查找：[comfyui-dynamicprompts](https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fcomfyui-dynamicprompts)。\n\n\u003Cp align=\"center\">\n    \u003Ca href=\"docs\u002FSYNTAX.md\">\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_5c08bc0775c8.png\" valign=\"middle\" style=\"height:60px;margin-right:10px\"\u002F>\u003C\u002Fa>\n    \u003Ca href=\"docs\u002Ftutorial.md\">\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_c9daf1469ba1.png\" valign=\"middle\" style=\"height:60px;margin-right:10px\"\u002F>\u003C\u002Fa>\n    \u003Ca href=\"docs\u002FCHANGELOG.md\">\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_a1a23345b43f.png\" valign=\"middle\" style=\"height:60px\"\u002F>\u003C\u002Fa>\n\u003C\u002Fp>\n\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_fed9dfd22d46.png\"\u002F>\n\n## 目录\n\n   * [基本用法](#basic-usage)\n   * [在线资源](#online-resources)\n   * [安装](#installation)\n   * [配置](#configuration)\n   * [故障排除](#troubleshooting)\n   * [兼容脚本](#compatible-scripts)\n   * [模板语法](#template-syntax)\n      * [模糊匹配\u002F递归通配符文件\u002F目录匹配](#fuzzy-globrecursive-wildcard-filedirectory-matching)\n   * [组合生成](#combinatorial-generation)\n      * [组合批次](#combinatorial-batches)\n      * [增加最大生成数量](#increasing-the-maximum-number-of-generations)\n   * [固定种子](#fixed-seed)\n   * [魔法提示词](#magic-prompt)\n      * [其他模型](#other-models)\n   * [让我来决定](#im-feeling-lucky)\n   * [吸引注意力](#attention-grabber)\n   * [将提示词写入文件](#write-prompts-to-file)\n   * [Jinja2 模板](#jinja2-templates)\n   * [WILDCARD_DIR](#wildcard_dir)\n   * [集合](#collections)\n   * [动态提示词与随机种子](#dynamic-prompts-and-random-seeds)\n      * [未启用动态提示词时](#without-dynamic-prompts-enabled)\n      * [在随机\u002F标准模式下启用动态提示词时](#with-dynamic-prompts-enabled-in-randomstandard-mode)\n      * [带有变体强度的动态提示词组合模式](#combinatorial-mode-with-variation-strength--0)\n\n\n## 基本用法\n使用此脚本，提示词：\n\n\tA {house|apartment|lodge|cottage} in {summer|winter|autumn|spring} by {2$$artist1|artist2|artist3}\n\n将会生成以下任意一种提示词：\n\n- 夏天里的**房子**，由**artist1**、**artist2**创作\n- 秋天里的**小屋**，由**artist3**、**artist1**创作\n- 冬天里的**别墅**，由**artist2**、**artist3**创作\n- ...\n\n这在您寻找有趣的艺术家和风格组合时尤其有用。\n\n您还可以从文件中随机选取一个字符串。假设您在 WILDCARD_DIR 中有 seasons.txt 文件（见下文），那么：\n\n    __seasons__ is coming\n\n可能会生成如下内容：\n\n- 冬天来了\n- 春天来了\n- ...\n\n您也可以两次使用同一个通配符：\n\n    I love __seasons__ better than __seasons__\n\n- 我更喜欢冬天而不是夏天\n- 我更喜欢春天而不是春天\n\n更多完整的文档可以在这里找到：[docs\u002FSYNTAX.md]。\u003Cbr\u002F>\n想要教程吗？\u003Ca href=\"docs\u002Ftutorial.md\">点击这里\u003C\u002Fa>\u003Cbr\u002F>\n需要通配符库？我们已经为您准备好了[集合](#collections)。\n\n## 在线资源\n您可以在这里找到教程和通配符包的列表：[docs\u002Fresources.md]\n\n## 安装\n\n该扩展可以直接从 Webui 的 **Extensions** 选项卡中安装。\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_4cb7c603b44c.png\"\u002F>\n\n您也可以手动安装，只需在 webui 目录下运行以下命令：\n\n```shell\ngit clone https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\u002F extensions\u002Fsd-dynamic-prompts\n```\n\n您可以在 `extensions\u002Fsd-dynamic-prompts\u002Fwildcards` 中创建[通配符文件](#template-syntax)，或者利用[预装的集合](#collections)，通过 Web UI 中的“Wildcards Manager”选项卡来使用。\n\n## 配置\n您可以在 Dynamic Prompts 部分的 Settings 选项卡中找到各种设置，以调整动态提示词的行为。\n\n### 更改语法\n如果与其他扩展发生语法冲突，Dynamic Prompts 允许您更改变体开始和结束的符号。默认情况下，它们分别设置为 `{` 和 `}`，例如 `{red|green|blue}`。在设置选项卡中，您可以将这两个符号更改为任何字符串，比如 `\u003Cred|green|blue>` 或者 `::red|green|blue::`。\n\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_8edfeea57f1c.png\">\n\n默认情况下，通配符以双下划线 `__` 开始并以 `__` 结束。您可以在设置选项卡中的 wildcard wrap 部分进行更改。\n\u003Cbr\u002F>\u003Cbr\u002F>\n\n### 通配符设置\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_6d498d91d25b.png\">\n\nDynamic Prompts 会在使用通配符文件之前自动去重并排序。如果您希望禁用此功能，可以在设置选项卡中取消勾选相关复选框。\n\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_4d7882be46ab.png>\n\n勾选“shuffle wildcards”复选框会随机化通配符的顺序，从而确保每次运行组合模型时都会生成不同的图像。\n\n### 将模板保存到元数据\n默认行为是将所有通配符解析为可用的提示词，然后将该提示词存储在 PNG 信息或 txt 文件中（例如 `I love Winter better than Summer`）。如果您还想保留包含通配符的原始模板，可以开启此选项，这样就会保存：\n\n```\n我更喜欢冬天而不是夏天\n模板：I love __seasons__ better than __seasons__\n```\n\n> 注意：这个额外的“Template”部分不会显示在生成页面上，但会在 PNG Info（或如果您安装了相应扩展的图片浏览器）中可见。\n\n## 故障排除\n如果您在使用 Dynamic Prompts 时遇到问题，请按照以下步骤进行排查和解决：\n\n1. 确保您已安装最新版本的 Dynamic Prompts 扩展和库。要检查已安装的版本，可在 txt2image 的 Dynamic Prompts 部分打开 **Need Help? 折叠面板**。扩展的最新版本号可在此处查看：[扩展](https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\u002Fblob\u002Fmain\u002Fdocs\u002FCHANGELOG.md)，库的最新版本号可在此处查看：[库](https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fdynamicprompts\u002Fblob\u002Fmain\u002FCHANGELOG.md?plain=1)。\n\n2. 如果版本不匹配，请在 Extensions（扩展）选项卡中更新扩展，并重启 WebUI。扩展应会自动更新库。\n\n3. 如果上述步骤无效，您可能需要手动运行以下命令来更新库：\n\n```shell\npython -m pip install -U dynamicprompts[attentiongrabber,magicprompt]\n```\n\n4. 重启 WebUI 并再次检查。如果 WebUI 使用的是不同的 Python 解释器，请找到正确的 Python 路径并运行：\n\n```shell\n\u002Fpath\u002Fto\u002Fpython\u002Fbinary\u002Fpython -m pip install -U dynamicprompts[attentiongrabber,magicprompt]\n```\n\n5. 如果 Wildcard UI 没有显示，可能是由于库版本过旧所致。请检查终端中的错误信息，并按照第 3 步所述更新库。\n\n6. 如果出现“未找到通配符 some\u002Fwildcard 的值”的错误提示，请确保文件 wildcard.txt 位于 extensions\u002Fsd-dynamic-prompts\u002Fwildcards\u002Fsome\u002F 目录下。必须提供完整路径，因为当前不支持相对路径。\n\n7. 如果问题仍然存在，请在 GitHub 的 [问题页面](https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\u002Fissues?q=is%3Aissue) 和 [讨论区](https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\u002Fdiscussions) 中查找解决方案。如果找不到解决方案，请创建一个新的问题，并为其起一个描述性的标题，例如“通配符值在提示模板中被忽略”。同时提供必要的上下文信息，包括 Dynamic Prompts 扩展和库的版本号，以及所使用的操作系统或 Colab 环境。如果终端中有错误信息，请复制粘贴全部内容或截屏。\n\n8. 最后，务必测试并应用我们发布的任何修复方案。您的反馈非常宝贵，因为在我们的环境中可以正常工作的问题，在您的环境中可能无法解决。\n\n## 兼容脚本\nDynamic Prompts 与 X\u002FY Plot 结合使用效果尤为出色——将 Dynamic Prompts 设置为 \u003Ca href=\"#combinatorial-generation\">组合模式\u003C\u002Fa> 同时使用 X\u002FY Plot，可以让您同时全面测试提示和参数的各种变化。\n\n## 模板语法\n文档可在 [这里](docs\u002FSYNTAX.md) 找到。\n\n### 模糊匹配与递归通配符文件\u002F目录匹配\n除了标准的通配符标记（如 `__times__` -> `times.txt`），您还可以使用 glob 匹配一次匹配多个文件。\n例如：\n\n> `__colors*__` 将匹配以下文件：\n> - WILDCARD_DIR\u002Fcolors.txt\n> - WILDCARD_DIR\u002Fcolors1.txt\n> - WILDCARD_DIR\u002Fnested\u002Ffolder\u002Fcolors1.txt\n>\n> `__light\u002F**\u002F*__` 将匹配：\n> - WILDCARD_DIR\u002Fnested\u002Ffolder\u002Flight\u002Fa.txt\n> - WILDCARD_DIR\u002Fnested\u002Ffolder\u002Flight\u002Fb.txt\n>\n> 但不会匹配：\n> - WILDCARD_DIR\u002Fnested\u002Ffolder\u002Fdark\u002Fa.txt\n> - WILDCARD_DIR\u002Fa.txt\n\n您还可以使用字符范围 `[0-9]` 和 `[a-z]`，以及单个通配符 `?`。更多示例请参阅 [这篇文章](http:\u002F\u002Fpymotw.com\u002F2\u002Fglob\u002F)。\n\n`WILDCARD_DIR` 默认为 `extensions\u002Fsd-dynamic-prompts\u002Fwildcards`。\n\n## 组合生成\n与从模板中随机生成提示不同，组合生成会根据给定的字符串生成所有可能的提示。例如：\n`I {love|hate} {New York|Chicago} in {June|July|August}`\n\n将生成：\n- I love New York in June\n- I love New York in July\n- I love New York in August\n- I love Chicago in June\n- I love Chicago in July\n- I love Chicago in August\n- I hate New York in June\n- I hate New York in July\n- I hate New York in August\n- I hate Chicago in June\n- I hate Chicago in July\n- I hate Chicago in August\n\n如果提供了 `__wildcard__`，则会为通配符文件中的每个值生成一个新的提示。例如：\n`My favourite season is __seasons__`\n\n将生成：\n- My favourite season is Summer\n- My favourite season is August\n- My favourite season is Winter\n- My favourite season is Sprint\n\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_0e402ecc982d.png\"\u002F>\n\n您还可以在通配符中嵌套组合，或在组合中嵌套通配符。\n\n组合生成在您希望为文件中的每位艺术家都生成一张图像时非常有用。可以通过勾选界面中的 __Combinatorial generation__ 复选框来启用此功能。为防止意外生成数千张图片，您可以使用 **Max Generations** 滑块限制生成的提示总数。默认值为 0，表示不限制。\n\n### 组合批次\n组合批次滑块允许您使用不同的种子重复同一组提示多次。默认批次数为 1。\n\n### 增加最大生成数量\n默认情况下，automatic1111 的 __Batch count__ 滑块的最大值为 100。这可能会限制使用组合生成时的最大生成数量。您可以通过编辑 ui-config.json 文件来更改此滑块的最大值，将：\n\ntxt2img\u002FBatch count\u002Fmaximum\": 100\n\n修改为更大的数值，例如：\n\ntxt2img\u002FBatch count\u002Fmaximum\": 1000\n\n## 固定种子\n如果您希望为每张生成的图片使用相同的种子，请选择此项。如果没有通配符，所有图片都将完全相同。这在您想要测试特定修饰符的效果时非常有用。例如：\n\nA beautiful day at the beach __medium\u002Fphotography\u002Ffilmtypes__\n\n这样您就可以单独分析每种胶片类型对特定场景的影响。以下是部分结果：\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_bd69d181a1bd.jpg\"\u002F>\n\n## Magic Prompt\n使用多种提示生成模型之一来丰富您的提示内容。\n\n借助 [Gustavosta](https:\u002F\u002Fhuggingface.co\u002FGustavosta\u002FMagicPrompt-Stable-Diffusion) 的 MagicPrompt 模型，该模型基于来自 [Lexica.art](lexica.art) 的 80,000 条提示进行训练，可以帮助您针对特定主题生成有趣的新提示。以下是关于“狗踢足球”的一些自动生成变体：\n\n> 狗踢足球，在日本小镇的夜晚街道上，人们惊奇地围观，吉卜力工作室和新海诚风格，高度精细的数字艺术，ArtStation 上热门作品\n\n> 狗踢足球，背景是一场核爆炸。照片级真实感。高清。超写实。4K。获奖作品。\n\n> 狗踢足球，背景是一场核爆炸。写实主义。4K 宽画幅。电影级。虚幻引擎。Artgerm。Marc Simonetti。JC Leyendecker\n\n此功能与上述通配符语法兼容。\n\n### 其他模型\n* [daspartho\u002Fprompt-extend (~500mb)](https:\u002F\u002Fhuggingface.co\u002Fdaspartho\u002Fprompt-extend)\n* [succinctly\u002Ftext2image-prompt-generator (~600mb)](https:\u002F\u002Fhuggingface.co\u002Fsuccinctly\u002Ftext2image-prompt-generator) - 基于Midjourney提示词训练\n* [microsoft\u002FPromptist (~500mb)](https:\u002F\u002Fhuggingface.co\u002Fmicrosoft\u002FPromptist) - 论文链接：[here](https:\u002F\u002Farxiv.org\u002Fabs\u002F2212.09611)\n* [AUTOMATIC\u002Fpromptgen-lexart (~300mb)](https:\u002F\u002Fhuggingface.co\u002FAUTOMATIC\u002Fpromptgen-lexart) - 使用lexica.art上的134,819条提示词微调\n* [AUTOMATIC\u002Fpromptgen-majinai-safe (~300mb)](https:\u002F\u002Fhuggingface.co\u002FAUTOMATIC\u002Fpromptgen-majinai-safe) - 来自majinai.art的1,654条提示词\n* [AUTOMATIC\u002Fpromptgen-majinai-unsafe (~300mb)](https:\u002F\u002Fhuggingface.co\u002FAUTOMATIC\u002Fpromptgen-majinai-unsafe) - 来自majinai.art的825条提示词（含NSFW内容）\n* [Gustavosta\u002FMagicPrompt-Dalle](https:\u002F\u002Fhuggingface.co\u002FGustavosta\u002FMagicPrompt-Dalle)\n* [kmewhort\u002Fstable-diffusion-prompt-bolster (~500mb)](https:\u002F\u002Fhuggingface.co\u002Fkmewhort\u002Fstable-diffusion-prompt-bolster),\n* [Ar4ikov\u002Fgpt2-650k-stable-diffusion-prompt-generator (~500mb)](Ar4ikov\u002Fgpt2-650k-stable-diffusion-prompt-generator),\n* [Ar4ikov\u002Fgpt2-medium-650k-stable-diffusion-prompt-generator (~1.4gb)](https:\u002F\u002Fhuggingface.co\u002FAr4ikov\u002Fgpt2-medium-650k-stable-diffusion-prompt-generator),\n* [crumb\u002Fbloom-560m-RLHF-SD2-prompter-aesthetic (~1.1gb)](https:\u002F\u002Fhuggingface.co\u002Fcrumb\u002Fbloom-560m-RLHF-SD2-prompter-aesthetic),\n* [Meli\u002FGPT2-Prompt (~500mb)](https:\u002F\u002Fhuggingface.co\u002FMeli\u002FGPT2-Prompt),\n* [DrishtiSharma\u002FStableDiffusion-Prompt-Generator-GPT-Neo-125M (~550mb)](https:\u002F\u002Fhuggingface.co\u002FDrishtiSharma\u002FStableDiffusion-Prompt-Generator-GPT-Neo-125M)\n\n首次使用某个模型时，它会被下载。该模型大约500MB，因此下载时间取决于您的网络速度。首次激活时，由于模型需要加载到内存中，也会花费几秒钟。请注意，如果您的显存不足，可能会出现CUDA错误。我的显卡使用不到8GB显存，但具体情况因人而异。\n\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_28f803324a6b.png\"\u002F>\n\n\n您可以通过__最大魔法提示长度__滑块来控制提示的最大长度。__魔法提示创意度__可以调整生成的提示，但您可能需要多次尝试以找到合适的设置。\n\n使用__魔法提示黑名单正则表达式__来过滤掉某些关键词。例如，如果您想避免包含Greg Rutkowski的提示，只需将他的名字添加到此字段中。\n\n如果您正在使用魔法提示生成大量提示，那么增加__魔法提示批次大小__可以显著提高提示生成速度。不过，这种提升只有在您不同时生成图像时才会明显，因为图像生成的速度远慢于提示生成。\n\n## 我很幸运\n使用[lexica.art](https:\u002F\u002Flexica.art) API来创建随机提示。这在您需要灵感或懒得自己思考提示时非常有用。当选择此选项时，主提示框中的内容将被用作搜索字符串。例如，提示“机械战士”可能会返回：\n\n* 由Greg Rutkowski、Sung Choi、Mitchell Mohrhauser、Maciej Kuciara、Johnson Ting、Maxim Verehin、Peter Konig创作的一座位于森林中央的巨大机器人石像，风格为最终幻想，8K超写实，电影级灯光，高清，细节丰富，氛围感十足。\n* 一幅精美的肖像画，描绘了一位身着赛博朋克装甲的人物，由simon stalenhag、pascal blanche、alphonse mucha和nekro共同创作，采用数字艺术风格。色彩鲜艳的漫画风，黑色电影般的对称构图，笔触细腻，色彩跃动，细节极其丰富。Octane渲染。在ArtStation上备受关注。\n* 对称！一位机器人宇航员的肖像，繁花似锦！出自《地平线零之曙光》中的机械设计，复杂精致、优雅且高度细节化，数字绘画，来自ArtStation的概念艺术作品，画面流畅，焦点清晰，插画风格，由artgerm、greg rutkowski和alphonse mucha等艺术家创作，8K分辨率。\n\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_761cfcf6fc59.png\">\n\n如果提示框留空，则会返回完全随机选择的提示列表。\n\n## 引人注目\n此选项会随机选择您提示中的一个关键词，并为其添加不同程度的强调。以下是其对提示的影响示例：\n\n一幅拟人化的熊猫法师施法的肖像，身穿法师长袍，背景是风景，可爱，DND角色艺术肖像，由Jason Felix和Peter Mohrbacher创作，电影级灯光。\n\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_ee751c9a1162.png\">\n\n勾选__高级选项__下的__固定种子__复选框，即可查看强调如何改变您的图像，而不会改变种子。\n\n## 将提示保存到文件\n勾选“将提示保存到文件”复选框，即可创建一个包含所有生成提示的文件。生成的文件是以提示内容为基础生成的URL友好字符串，可与生成的图片存放在同一目录下，例如：outputs\u002Ftxt2img-images。\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_8dfe6d623df5.png\"\u002F>\n\n## Jinja2模板\n[Jinja2模板](https:\u002F\u002Fjinja.palletsprojects.com\u002Fen\u002F3.1.x\u002Ftemplates\u002F)是一项实验性功能，允许您以编程方式定义提示。这是一项高级功能，仅推荐给熟悉脚本编写的用户。\n\n要启用，请打开“高级”折叠面板并选择“启用Jinja2模板”。\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_readme_37b92400c8a9.png\">\n\n您可以点击\u003Ca href=\"jinja2.md\">这里\u003C\u002Fa>了解更多详细信息。\n\n## WILDCARD_DIR\n该扩展会在WILDCARD_DIR目录中查找通配符文件。默认路径为`\u002Fpath\u002Fto\u002Fstable-diffusion-webui\u002Fextensions\u002Fsd-dynamic-prompts\u002Fwildcards`。您也可以在WebUI的主配置文件config.json中手动指定wildcard_dir。如有疑问，可在WebUI扩展的帮助文本中找到WILDCARD_DIR的完整路径。\n\n## 资源库\ncollections 目录包含修饰符库，您可以直接使用，也可以在此基础上构建自己的资源库。要开始使用，您可以通过“通配符管理器”选项卡将一个或多个资源库复制到您的 wildcards 文件夹中，或者手动复制文件。动态提示扩展程序自带三个资源库。\n\n- [jumbo](https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\u002Ftree\u002Fmain\u002Fcollections\u002Fjumbo)\n  这是一个非常庞大的通配符资源库，涵盖了美学、外观、艺术家、媒介、风格和时间等多个类别。该资源库仍在不断完善中，旨在全面覆盖各类修饰符类别。\n  \n- [parrotzone](https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\u002Ftree\u002Fmain\u002Fcollections\u002Fparrotzone)\n  这是一个规模较小、更易于管理的资源库，内容来源于 https:\u002F\u002Fproximacentaurib.notion.site\u002Fe28a4f8d97724f14a784a538b8589e7d?v=42948fd8f45c4d47a0edfc4b78937474。\n  \n- [devilkkw](https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\u002Ftree\u002Fmain\u002Fcollections\u002Fdevilkkw)\n  Devilkkw 更侧重于角色塑造、服饰、手势、食物等内容。\n\n如果您使用的是 Unix\u002FLinux 操作系统，并且不打算修改相关资源库，可以直接创建指向相应资源库的符号链接，而无需将其复制过去。例如：\n\n```shell\nln -sr collections\u002Fparrotzone wildcards\u002F\n```\n\n您还可以通过在扩展程序的根目录（即 `extensions\u002Fsd-dynamic-prompts\u002F`）下运行 `python _tools\u002Fdownload_collections.py` 来下载其他扩展资源库。\n\n## 动态提示与随机种子\n随机种子在控制生成结果的随机性方面起着重要作用。下面我们来探讨在不同场景下，动态提示如何与随机种子协同工作。\n\n### 未启用动态提示时\n\n1. 如果种子设置为 -1：系统会随机选择一个种子。该种子用于生成第一张图像，随后每张图像都会在前一张图像的种子基础上加 1 来生成，以此类推。\n2. 如果种子设置为大于 -1 的具体数值：过程与上述类似，但会从用户指定的种子开始。\n3. 如果定义了变化种子，但变化强度为零：过程与前两点相同。\n4. 如果变化种子设置为大于 0 的数值：每张图像都使用相同的初始种子（随机选择或用户指定）。变化种子可以是随机的（当设置为 -1 时）或由用户指定的值。第一张图像使用变化种子生成，第二张则使用变化种子 + 1，依此类推。\n\n### 启用动态提示并使用随机\u002F标准模式时\n\n1. 如果种子设置为 -1：过程与上一节的第一点相似。不过，此时提示也会根据相同的种子进行选择（如果使用随机提示生成器）。\n2. 如果种子设置为大于 -1 的数值：过程与上一节的第二点相似。区别在于，此时也会根据选定的种子生成随机提示（如果使用提示生成器）。\n3. 如果勾选了固定种子复选框：所有图像和提示都将使用同一个种子。这意味着会反复生成同一张图像（这在组合式生成中很有用）。\n4. 如果同时勾选了固定种子和取消提示与种子关联的复选框：提示将使用随机种子，但所有图像仍会使用同一个种子。如果您想了解不同的提示对同一张图像生成效果的影响，这种设置可能会有所帮助。\n\n### 带有动态提示的变化种子\n\n1. 变化强度设置为 0：变化将被忽略。\n2. 变化种子设置为大于 0 的数值：每张图像都会分配一个变化种子，每次递增 1。然而，由于您希望得到同一张图像的不同变体，因此只会生成 1 张提示。\n\n### 变化强度大于 0 的组合模式\n在这种情况下，系统只会为您生成第一张图像，这可能并不是您想要的结果。为了获得理想的效果，您可能需要调整设置或改用其他模式。","# sd-dynamic-prompts 快速上手指南\n\n`sd-dynamic-prompts` 是专为 AUTOMATIC1111 Stable Diffusion WebUI 设计的扩展插件。它提供了一套强大的模板语言，支持随机提示词生成、组合式批量生成以及深层通配符目录结构，帮助用户高效探索艺术家风格与提示词组合。\n\n## 环境准备\n\n*   **系统要求**：已安装并正常运行 [AUTOMATIC1111\u002Fstable-diffusion-webui](https:\u002F\u002Fgithub.com\u002FAUTOMATIC1111\u002Fstable-diffusion-webui) 的环境（Windows, Linux, macOS）。\n*   **前置依赖**：\n    *   Python 3.10+ (WebUI 自带环境即可)。\n    *   Git（用于克隆仓库）。\n    *   稳定的网络连接以访问 GitHub 或 HuggingFace（如需使用 Magic Prompt 功能）。\n\n## 安装步骤\n\n推荐通过 WebUI 界面直接安装，也可选择手动命令行安装。\n\n### 方法一：WebUI 界面安装（推荐）\n\n1.  启动 Stable Diffusion WebUI。\n2.  点击顶部导航栏的 **Extensions**（扩展）标签页。\n3.  选择 **Install from URL**（从网址安装）子标签。\n4.  在 \"URL for extension's git repository\" 输入框中填入：\n    ```text\n    https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\n    ```\n5.  点击 **Install** 按钮。\n6.  安装完成后，切换到 **Installed** 标签页，点击 **Apply and restart UI** 应用并重启界面。\n\n### 方法二：命令行手动安装\n\n进入 WebUI 根目录，执行以下命令：\n\n```shell\ngit clone https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts extensions\u002Fsd-dynamic-prompts\n```\n\n执行完毕后重启 WebUI。\n\n> **注意**：若安装后功能未生效或报错，可能需要更新底层库。请在 WebUI 对应的 Python 环境中运行：\n> ```shell\n> python -m pip install -U dynamicprompts[attentiongrabber,magicprompt]\n> ```\n> *(国内用户如遇 pip 下载缓慢，可添加清华源：`-i https:\u002F\u002Fpypi.tuna.tsinghua.edu.cn\u002Fsimple`)*\n\n## 基本使用\n\n安装重启后，在 `txt2img` 或 `img2img` 页面下方即可看到 **Dynamic Prompts** 相关选项。\n\n### 1. 随机变体生成 (Random Generation)\n\n使用 `{选项 A|选项 B|选项 C}` 语法，每次生成时随机选择一个选项。\n\n**示例提示词：**\n```text\nA {house|apartment|lodge|cottage} in {summer|winter|autumn|spring} by {2$$artist1|artist2|artist3}\n```\n\n**生成结果示例：**\n*   `A house in summer by artist1, artist2`\n*   `A lodge in autumn by artist3, artist1`\n*   `A cottage in winter by artist2, artist3`\n\n*注：`2$$` 表示从后面的列表中随机选取 2 个不重复的项。*\n\n### 2. 通配符文件引用 (Wildcards)\n\n你可以从文本文件中随机读取一行内容。默认通配符目录位于 `extensions\u002Fsd-dynamic-prompts\u002Fwildcards`。\n\n**步骤：**\n1.  在 `wildcards` 目录下创建文件 `seasons.txt`，内容如下：\n    ```text\n    Spring\n    Summer\n    Autumn\n    Winter\n    ```\n2.  在提示词中使用双下划线调用：\n    ```text\n    __seasons__ is coming\n    ```\n\n**生成结果示例：**\n*   `Winter is coming`\n*   `Spring is coming`\n\n你也可以在同一提示词中多次使用同一通配符：\n```text\nI love __seasons__ better than __seasons__\n```\n可能生成：`I love Winter better than Summer`\n\n### 3. 组合式批量生成 (Combinatorial Generation)\n\n如果你希望生成**所有可能**的组合而不是随机选择，请勾选界面上的 **Combinatorial generation** 复选框。\n\n**示例提示词：**\n```text\nI {love|hate} {New York|Chicago}\n```\n\n**将依次生成以下 4 张图：**\n1.  `I love New York`\n2.  `I love Chicago`\n3.  `I hate New York`\n4.  `I hate Chicago`\n\n*提示：可通过 **Max Generations** 滑块限制最大生成数量，防止意外生成过多图片。*","一位独立游戏开发者正在为奇幻 RPG 项目批量生成数百张风格各异的“季节与建筑”概念图，以丰富游戏场景库。\n\n### 没有 sd-dynamic-prompts 时\n- **手动组合效率极低**：开发者需反复手动修改提示词中的“房屋类型”和“季节”，如将\"house\"改为\"cottage\"、\"summer\"改为\"winter\"，耗时且枯燥。\n- **创意覆盖范围狭窄**：受限于人工操作的耐心，只能测试寥寥几种固定搭配，难以发现“秋季小屋配特定画家风格”等意外惊艳的组合。\n- **外部资源调用繁琐**：若想从本地文本文件随机读取大量艺术家名字或材质描述，必须编写额外的 Python 脚本或手动复制粘贴，无法在 WebUI 内直接完成。\n- **批次生成缺乏变化**：一次性生成多张图片时，若不开启复杂设置，往往得到的是完全相同的提示词结果，失去了随机探索的意义。\n\n### 使用 sd-dynamic-prompts 后\n- **模板化自动排列组合**：只需输入 `A {house|cottage|lodge} in {__seasons__} by {2$$artist}`，sd-dynamic-prompts 即可自动穷举或随机生成成千上万种不重复的提示词组合。\n- **深度挖掘创意空间**：通过通配符语法轻松调用本地 `seasons.txt` 或预置艺术家库，让 AI 自动尝试人类难以想到的冷门风格混搭，激发设计灵感。\n- **原生支持文件随机读取**：直接在提示词中使用 `__filename__` 语法即可动态加载外部文本内容，无需任何额外编程，工作流无缝集成在 AUTOMATIC1111 界面中。\n- **高效批量生产变体**：配合组合模式（Combinatorial Mode），一次点击即可基于单一模板生成涵盖所有变量组合的图像批次，极大提升了素材产出速度。\n\nsd-dynamic-prompts 将原本机械重复的提示词编写工作转化为高效的自动化创意实验，让开发者能专注于筛选最佳视觉方案而非陷入文字堆砌。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fadieyal_sd-dynamic-prompts_5c08bc07.png","adieyal","Adi Eyal","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Fadieyal_45300913.png",null,"https:\u002F\u002Fgithub.com\u002Fadieyal",[79,83,87,91],{"name":80,"color":81,"percentage":82},"Python","#3572A5",80.8,{"name":84,"color":85,"percentage":86},"JavaScript","#f1e05a",13.2,{"name":88,"color":89,"percentage":90},"HTML","#e34c26",4.3,{"name":92,"color":93,"percentage":94},"CSS","#663399",1.7,2260,295,"2026-04-08T07:25:54","MIT","未说明",{"notes":101,"python":99,"dependencies":102},"该工具是 AUTOMATIC1111 Stable Diffusion WebUI 的扩展插件，并非独立运行的程序。因此其运行环境完全依赖于宿主 WebUI 的环境配置（通常包括 Python 3.10+、PyTorch 及相应的 GPU 驱动）。安装时可通过 WebUI 界面直接安装或手动 git clone，部分功能（如 Magic Prompt）需要额外安装 dynamicprompts 库及其可选依赖。",[21,103],"dynamicprompts",[15,43],"2026-03-27T02:49:30.150509","2026-04-09T01:07:12.969668",[108,113,118,123,128,133],{"id":109,"question_zh":110,"answer_zh":111,"source_url":112},25339,"启用动态提示（Dynamic Prompts）后，生成的提示词被截断或只显示部分内容怎么办？","这通常是由于版本更新导致的空格处理问题。请尝试将扩展升级到 v2.1.0 或更高版本，该版本专门修复了相关的空格和截断问题。如果问题依旧，请检查提示词结构，确保使用正确的语法（如 [word1|word2]），并重启 WebUI 界面。","https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\u002Fissues\u002F144",{"id":114,"question_zh":115,"answer_zh":116,"source_url":117},25340,"为什么开启“魔法提示”（Magic Prompt）后，额外的提示词被添加到了负面提示词（Negative Prompt）中而不是正面提示词？","这通常是由与其他扩展（特别是 \"Unprompted\" 扩展）冲突引起的。这两个扩展在提示词处理功能上有重叠，不建议同时运行。解决方法是禁用 \"Unprompted\" 扩展，或者只保留其中一个扩展处于激活状态。可以通过逐个启用扩展的方法来排查具体的冲突源。","https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\u002Fissues\u002F62",{"id":119,"question_zh":120,"answer_zh":121,"source_url":122},25341,"组合生成（Combinatorial Generation）在使用多个通配符时失效或导致程序崩溃怎么办？","这是早期版本的已知限制。由于该扩展已历经多次重大更新，旧版中关于多通配符导致静默失败的问题可能已被修复或改变。如果当前版本仍遇到此问题，建议不要在一个模板中使用过多的嵌套通配符，或者提交新的 Issue 并提供复现步骤，因为旧的解决方案可能已不再适用。","https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\u002Fissues\u002F27",{"id":124,"question_zh":125,"answer_zh":126,"source_url":127},25342,"安装扩展后，动态提示功能完全不起作用（输入 __wildcard__ 或 {} 语法无反应）如何解决？","此问题通常与特定的环境配置或重复的 Issue #338 有关。请查阅 Issue #338 中的修复说明。常见的解决步骤包括：确认 Wildcards 文件夹路径正确、检查 WebUI 控制台日志是否有报错、尝试重新安装扩展或回退到稳定的 WebUI 版本（如 1.5）。如果确认是重复问题，应用该线程中提供的修复方案通常能解决问题。","https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\u002Fissues\u002F344",{"id":129,"question_zh":130,"answer_zh":131,"source_url":132},25343,"运行扩展时出现 \"Error running process\" 或 Python  traceback 错误（如 prompt parser issues）怎么办？","这类错误通常源于提示词解析失败或脚本执行异常。首先检查您的提示词语法是否符合规范（例如括号是否闭合、通配符格式是否正确）。其次，查看完整的错误日志（Traceback），定位出错的具体文件行号。如果是由于特定提示词模板引起的，尝试简化模板以隔离问题。确保您使用的是与当前 Stable Diffusion WebUI 版本兼容的扩展版本。","https:\u002F\u002Fgithub.com\u002Fadieyal\u002Fsd-dynamic-prompts\u002Fissues\u002F142",{"id":134,"question_zh":135,"answer_zh":136,"source_url":122},25344,"组合生成模式下，提示词的生成顺序为什么不一致且每次重启后都会变化？","这是因为扩展在组合生成中引入了随机洗牌（shuffle）机制。当生成的提示词总数超过“批次大小 × 批次数量”时，系统会随机选择一个子集，导致顺序不固定。此外，提示词文本的微小变化（如多余的空格）也可能影响哈希值从而导致顺序改变。如果需要固定的顺序，目前可能需要避免触发随机子集逻辑，或者在代码层面调整相关生成器设置（如 `combinatorial.py` 中的逻辑）。",[]]