[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-dezoito--ollama-grid-search":3,"tool-dezoito--ollama-grid-search":64},[4,17,27,35,43,56],{"id":5,"name":6,"github_repo":7,"description_zh":8,"stars":9,"difficulty_score":10,"last_commit_at":11,"category_tags":12,"status":16},3808,"stable-diffusion-webui","AUTOMATIC1111\u002Fstable-diffusion-webui","stable-diffusion-webui 是一个基于 Gradio 构建的网页版操作界面，旨在让用户能够轻松地在本地运行和使用强大的 Stable Diffusion 图像生成模型。它解决了原始模型依赖命令行、操作门槛高且功能分散的痛点，将复杂的 AI 绘图流程整合进一个直观易用的图形化平台。\n\n无论是希望快速上手的普通创作者、需要精细控制画面细节的设计师，还是想要深入探索模型潜力的开发者与研究人员，都能从中获益。其核心亮点在于极高的功能丰富度：不仅支持文生图、图生图、局部重绘（Inpainting）和外绘（Outpainting）等基础模式，还独创了注意力机制调整、提示词矩阵、负向提示词以及“高清修复”等高级功能。此外，它内置了 GFPGAN 和 CodeFormer 等人脸修复工具，支持多种神经网络放大算法，并允许用户通过插件系统无限扩展能力。即使是显存有限的设备，stable-diffusion-webui 也提供了相应的优化选项，让高质量的 AI 艺术创作变得触手可及。",162132,3,"2026-04-05T11:01:52",[13,14,15],"开发框架","图像","Agent","ready",{"id":18,"name":19,"github_repo":20,"description_zh":21,"stars":22,"difficulty_score":23,"last_commit_at":24,"category_tags":25,"status":16},1381,"everything-claude-code","affaan-m\u002Feverything-claude-code","everything-claude-code 是一套专为 AI 编程助手（如 Claude Code、Codex、Cursor 等）打造的高性能优化系统。它不仅仅是一组配置文件，而是一个经过长期实战打磨的完整框架，旨在解决 AI 代理在实际开发中面临的效率低下、记忆丢失、安全隐患及缺乏持续学习能力等核心痛点。\n\n通过引入技能模块化、直觉增强、记忆持久化机制以及内置的安全扫描功能，everything-claude-code 能显著提升 AI 在复杂任务中的表现，帮助开发者构建更稳定、更智能的生产级 AI 代理。其独特的“研究优先”开发理念和针对 Token 消耗的优化策略，使得模型响应更快、成本更低，同时有效防御潜在的攻击向量。\n\n这套工具特别适合软件开发者、AI 研究人员以及希望深度定制 AI 工作流的技术团队使用。无论您是在构建大型代码库，还是需要 AI 协助进行安全审计与自动化测试，everything-claude-code 都能提供强大的底层支持。作为一个曾荣获 Anthropic 黑客大奖的开源项目，它融合了多语言支持与丰富的实战钩子（hooks），让 AI 真正成长为懂上",138956,2,"2026-04-05T11:33:21",[13,15,26],"语言模型",{"id":28,"name":29,"github_repo":30,"description_zh":31,"stars":32,"difficulty_score":23,"last_commit_at":33,"category_tags":34,"status":16},2271,"ComfyUI","Comfy-Org\u002FComfyUI","ComfyUI 是一款功能强大且高度模块化的视觉 AI 引擎，专为设计和执行复杂的 Stable Diffusion 图像生成流程而打造。它摒弃了传统的代码编写模式，采用直观的节点式流程图界面，让用户通过连接不同的功能模块即可构建个性化的生成管线。\n\n这一设计巧妙解决了高级 AI 绘图工作流配置复杂、灵活性不足的痛点。用户无需具备编程背景，也能自由组合模型、调整参数并实时预览效果，轻松实现从基础文生图到多步骤高清修复等各类复杂任务。ComfyUI 拥有极佳的兼容性，不仅支持 Windows、macOS 和 Linux 全平台，还广泛适配 NVIDIA、AMD、Intel 及苹果 Silicon 等多种硬件架构，并率先支持 SDXL、Flux、SD3 等前沿模型。\n\n无论是希望深入探索算法潜力的研究人员和开发者，还是追求极致创作自由度的设计师与资深 AI 绘画爱好者，ComfyUI 都能提供强大的支持。其独特的模块化架构允许社区不断扩展新功能，使其成为当前最灵活、生态最丰富的开源扩散模型工具之一，帮助用户将创意高效转化为现实。",107662,"2026-04-03T11:11:01",[13,14,15],{"id":36,"name":37,"github_repo":38,"description_zh":39,"stars":40,"difficulty_score":23,"last_commit_at":41,"category_tags":42,"status":16},3704,"NextChat","ChatGPTNextWeb\u002FNextChat","NextChat 是一款轻量且极速的 AI 助手，旨在为用户提供流畅、跨平台的大模型交互体验。它完美解决了用户在多设备间切换时难以保持对话连续性，以及面对众多 AI 模型不知如何统一管理的痛点。无论是日常办公、学习辅助还是创意激发，NextChat 都能让用户随时随地通过网页、iOS、Android、Windows、MacOS 或 Linux 端无缝接入智能服务。\n\n这款工具非常适合普通用户、学生、职场人士以及需要私有化部署的企业团队使用。对于开发者而言，它也提供了便捷的自托管方案，支持一键部署到 Vercel 或 Zeabur 等平台。\n\nNextChat 的核心亮点在于其广泛的模型兼容性，原生支持 Claude、DeepSeek、GPT-4 及 Gemini Pro 等主流大模型，让用户在一个界面即可自由切换不同 AI 能力。此外，它还率先支持 MCP（Model Context Protocol）协议，增强了上下文处理能力。针对企业用户，NextChat 提供专业版解决方案，具备品牌定制、细粒度权限控制、内部知识库整合及安全审计等功能，满足公司对数据隐私和个性化管理的高标准要求。",87618,"2026-04-05T07:20:52",[13,26],{"id":44,"name":45,"github_repo":46,"description_zh":47,"stars":48,"difficulty_score":23,"last_commit_at":49,"category_tags":50,"status":16},2268,"ML-For-Beginners","microsoft\u002FML-For-Beginners","ML-For-Beginners 是由微软推出的一套系统化机器学习入门课程，旨在帮助零基础用户轻松掌握经典机器学习知识。这套课程将学习路径规划为 12 周，包含 26 节精炼课程和 52 道配套测验，内容涵盖从基础概念到实际应用的完整流程，有效解决了初学者面对庞大知识体系时无从下手、缺乏结构化指导的痛点。\n\n无论是希望转型的开发者、需要补充算法背景的研究人员，还是对人工智能充满好奇的普通爱好者，都能从中受益。课程不仅提供了清晰的理论讲解，还强调动手实践，让用户在循序渐进中建立扎实的技能基础。其独特的亮点在于强大的多语言支持，通过自动化机制提供了包括简体中文在内的 50 多种语言版本，极大地降低了全球不同背景用户的学习门槛。此外，项目采用开源协作模式，社区活跃且内容持续更新，确保学习者能获取前沿且准确的技术资讯。如果你正寻找一条清晰、友好且专业的机器学习入门之路，ML-For-Beginners 将是理想的起点。",84991,"2026-04-05T10:45:23",[14,51,52,53,15,54,26,13,55],"数据工具","视频","插件","其他","音频",{"id":57,"name":58,"github_repo":59,"description_zh":60,"stars":61,"difficulty_score":10,"last_commit_at":62,"category_tags":63,"status":16},3128,"ragflow","infiniflow\u002Fragflow","RAGFlow 是一款领先的开源检索增强生成（RAG）引擎，旨在为大语言模型构建更精准、可靠的上下文层。它巧妙地将前沿的 RAG 技术与智能体（Agent）能力相结合，不仅支持从各类文档中高效提取知识，还能让模型基于这些知识进行逻辑推理和任务执行。\n\n在大模型应用中，幻觉问题和知识滞后是常见痛点。RAGFlow 通过深度解析复杂文档结构（如表格、图表及混合排版），显著提升了信息检索的准确度，从而有效减少模型“胡编乱造”的现象，确保回答既有据可依又具备时效性。其内置的智能体机制更进一步，使系统不仅能回答问题，还能自主规划步骤解决复杂问题。\n\n这款工具特别适合开发者、企业技术团队以及 AI 研究人员使用。无论是希望快速搭建私有知识库问答系统，还是致力于探索大模型在垂直领域落地的创新者，都能从中受益。RAGFlow 提供了可视化的工作流编排界面和灵活的 API 接口，既降低了非算法背景用户的上手门槛，也满足了专业开发者对系统深度定制的需求。作为基于 Apache 2.0 协议开源的项目，它正成为连接通用大模型与行业专有知识之间的重要桥梁。",77062,"2026-04-04T04:44:48",[15,14,13,26,54],{"id":65,"github_repo":66,"name":67,"description_en":68,"description_zh":69,"ai_summary_zh":69,"readme_en":70,"readme_zh":71,"quickstart_zh":72,"use_case_zh":73,"hero_image_url":74,"owner_login":75,"owner_name":76,"owner_avatar_url":77,"owner_bio":78,"owner_company":79,"owner_location":79,"owner_email":79,"owner_twitter":79,"owner_website":80,"owner_url":81,"languages":82,"stars":103,"forks":104,"last_commit_at":105,"license":106,"difficulty_score":23,"env_os":107,"env_gpu":108,"env_ram":108,"env_deps":109,"category_tags":112,"github_topics":113,"view_count":10,"oss_zip_url":79,"oss_zip_packed_at":79,"status":16,"created_at":122,"updated_at":123,"faqs":124,"releases":155},1232,"dezoito\u002Follama-grid-search","ollama-grid-search","A multi-platform desktop application to evaluate and compare LLM models, written in Rust and React.","ollama-grid-search 是一款跨平台的桌面应用，旨在帮助用户快速评估和比较不同大型语言模型（LLM）在特定任务中的表现。它通过自动化测试多个模型、提示词和推理参数的组合，让用户直观地查看结果，从而找到最优配置。\n\n这款工具解决了在选择合适的模型或提示词时需要反复手动测试的问题，节省了大量时间和精力。用户可以同时对多个模型进行 A\u002FB 测试，并根据实验结果进行筛选和优化。\n\n它适合研究人员、开发者以及对 AI 模型性能有较高要求的用户使用，尤其适合那些希望高效探索不同模型和提示组合效果的人群。其独特的亮点包括支持自定义参数设置、可重复运行实验、内置提示库以及对并发请求的控制功能，使实验过程更加灵活和可控。\n\n此外，它基于 Rust 和 React 开发，具备良好的性能和用户体验，支持 macOS、Windows 和 Linux 系统，方便各类用户直接下载使用。","# Ollama Grid Search: Instantly Evaluate Multiple LLMs and Prompts.\n\nThis project automates the process of selecting the best models, prompts, or inference parameters for a given use-case, allowing you to iterate over their combinations and to visually inspect the results.\n\nIt assumes [Ollama](https:\u002F\u002Fwww.ollama.ai) is installed and serving endpoints, either in `localhost` or in a remote server.\n\nHere's what an experiment for a simple prompt, tested on 3 different models, looks like:\n\n[\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdezoito_ollama-grid-search_readme_4c07171c4df1.png\" alt=\"Main Screenshot\" width=\"720\">](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdezoito_ollama-grid-search_readme_4c07171c4df1.png)\n\n**Dowloads: · [macOS](https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Freleases\u002Flatest) · [Windows](https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Freleases\u002Flatest) · [Linux](https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Freleases\u002Flatest) · [All Releases](https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Freleases)**\n\n(For a more in-depth look at an evaluation process assisted by this tool, please check https:\u002F\u002Fdezoito.github.io\u002F2023\u002F12\u002F27\u002Frust-ollama-grid-search.html).\n\n## Table of Contents\n\n- [Installation](#installation)\n- [Features](#features)\n- [Grid Search Concept](#grid-search-or-something-similar)\n- [A\u002FB Testing](#ab-testing)\n- [Prompt Archive](#prompt-archive)\n- [Experiment Logs](#experiment-logs)\n- [Future Features](#future-features)\n- [Contributing](#contributing)\n- [Development](#development)\n- [Citations](#citations)\n- [Acknowledgements](#thank-you)\n\n## Installation\n\nCheck the [releases page](https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Freleases) for the project, or on the sidebar.\n\n### macOS Installation (Apple Silicon)\n\nFor Apple Silicon (M1\u002FM2\u002FM3\u002FM4) Macs, the application uses ad-hoc code signing (a free alternative to paid Apple Developer certificates). When first launching the app:\n\n1. Download the `.dmg` file for your architecture (aarch64 for Apple Silicon)\n2. Open the DMG and drag the app to Applications\n3. **Right-click** (or Control+click) the application and select **\"Open\"**\n4. Click **\"Open\"** in the security dialog that appears\n5. The app will now launch normally (subsequent launches can use double-click)\n\nThis is necessary because the app is not notarized with a paid Apple Developer account. The app is safe to use - this is standard for open-source macOS applications distributed outside the App Store.\n\n## Features\n\n- Automatically fetches models from local or remote Ollama servers;\n- Iterates over multiple different models, prompts and parameters to generate inferences;\n- A\u002FB test different prompts on several models simultaneously;\n- Allows multiple iterations for each combination of parameters;\n- Allows [limited concurrency](https:\u002F\u002Fdezoito.github.io\u002F2024\u002F03\u002F21\u002Freact-limited-concurrency.html) **or** synchronous inference calls (to prevent spamming servers);\n- Optionally outputs inference parameters and response metadata (inference time, tokens and tokens\u002Fs);\n- Refetching of individual inference calls;\n- Model selection can be filtered by name;\n- List experiments which can be downloaded in JSON format;\n- Experiments can be inspected in readable views;\n- Re-run past experiments, cloning or modifying the parameters used in the past;\n- Configurable inference timeout;\n- Custom default parameters and system prompts can be defined in settings\n- Fully functional prompt database with examples;\n- Prompts can be selected and \"autocompleted\" by typing \"\u002F\" in the inputs\n\n## Grid Search (or something similar...)\n\nTechnically, the term \"grid search\" refers to iterating over a series of different model hyperparams to optimize model performance, but that usually means parameters like `batch_size`, `learning_rate`, or `number_of_epochs`, more commonly used in training.\n\nBut the concept here is similar:\n\nLets define a selection of models, a prompt and some parameter combinations:\n\n[\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdezoito_ollama-grid-search_readme_deb35749a427.gif\" alt=\"gridparams\" width=\"400\">](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdezoito_ollama-grid-search_readme_deb35749a427.gif)\n\nThe prompt will be submitted once for each parameter **value**, for each one of the selected models, generating a set of responses.\n\n## A\u002FB Testing\n\nSimilarly, you can perform A\u002FB tests by selecting different models and compare results for the same prompt\u002Fparameter combination, or test different prompts under similar configurations:\n\n[\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdezoito_ollama-grid-search_readme_cb21de788495.gif\" alt=\"A\u002FB testing\" width=\"720\">](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdezoito_ollama-grid-search_readme_cb21de788495.gif)\n\n\u003Csmall>Comparing the results of different prompts for the same model\u003C\u002Fsmall>\n\n## Prompt Archive\n\nYou can save and manage your prompts (we want to make prompts compatible with [Open WebUI](https:\u002F\u002Fgithub.com\u002Fopen-webui\u002Fopen-webui))\n\n[\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdezoito_ollama-grid-search_readme_0ffe38c6b4be.png\" alt=\"Settings\" width=\"720\">](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdezoito_ollama-grid-search_readme_0ffe38c6b4be.png)\n\nYou can **autocomplete** prompts by typing \"\u002F\" (inspired by Open WebUI, as well):\n\n[\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdezoito_ollama-grid-search_readme_827153289fd3.gif\" alt=\"A\u002FB testing\" width=\"720\">](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdezoito_ollama-grid-search_readme_827153289fd3.gif)\n\n## Experiment Logs\n\nYou can list, inspect, or download your experiments:\n\n[\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdezoito_ollama-grid-search_readme_7137902746a3.png\" alt=\"Settings\" width=\"720\">](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdezoito_ollama-grid-search_readme_7137902746a3.png)\n\n## Future Features\n\n- Grading results and filtering by grade\n- Importing, exporting and sharing prompt lists and experiment files.\n\n## Contributing\n\n- For obvious bugs and spelling mistakes, please go ahead and submit a PR.\n\n- If you want to propose a new feature, change existing functionality, or propose anything more complex, please open an issue for discussion, **before** getting work done on a PR.\n\n## Development\n\nThe [development notes](.\u002Fdocs\u002FDEVELOPMENT.md) provide setup instructions, sequence diagrams, and workflow charts that should make it easier to understand the project and get started.\n\n## Citations\n\nThe following works and theses have cited this repository:\n\nInouye, D & Lindo, L, & Lee, R & Allen, E; Computer Science and Engineering Senior Theses: **Applied Auto-tuning on LoRA Hyperparameters**\nSanta Clara University, 2024\n\u003Chttps:\u002F\u002Fscholarcommons.scu.edu\u002Fcgi\u002Fviewcontent.cgi?article=1271&context=cseng_senior>\n\n## Thank you!\n\nHuge thanks to [@FabianLars](https:\u002F\u002Fgithub.com\u002FFabianLars), [@peperroni21](https:\u002F\u002Fgithub.com\u002Fpepperoni21) and [@TomReidNZ](https:\u002F\u002Fgithub.com\u002FTomReidNZ).\n","# Ollama 网格搜索：即时评估多个大语言模型与提示\n\n本项目旨在自动化为特定用例选择最佳模型、提示或推理参数的过程，允许您遍历这些组合并直观地查看结果。\n\n本项目假设已安装 [Ollama](https:\u002F\u002Fwww.ollama.ai) 并在本地或远程服务器上提供服务端点。\n\n以下是一个针对简单提示、在 3 种不同模型上进行测试的实验示例：\n\n![主界面截图](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdezoito_ollama-grid-search_readme_4c07171c4df1.png)\n\n**下载：· [macOS](https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Freleases\u002Flatest) · [Windows](https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Freleases\u002Flatest) · [Linux](https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Freleases\u002Flatest) · [所有版本](https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Freleases)**\n\n（如需深入了解该工具辅助的评估流程，请参阅 https:\u002F\u002Fdezoito.github.io\u002F2023\u002F12\u002F27\u002Frust-ollama-grid-search.html。）\n\n## 目录\n\n- [安装](#installation)\n- [功能](#features)\n- [网格搜索概念](#grid-search-or-something-similar)\n- [A\u002FB 测试](#ab-testing)\n- [提示档案](#prompt-archive)\n- [实验日志](#experiment-logs)\n- [未来功能](#future-features)\n- [贡献](#contributing)\n- [开发](#development)\n- [引用](#citations)\n- [致谢](#thank-you)\n\n## 安装\n\n请在 [发布页面](https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Freleases) 或侧边栏中查看该项目。\n\n### macOS 安装（Apple Silicon）\n\n对于 Apple Silicon（M1\u002FM2\u002FM3\u002FM4）Mac，应用程序采用临时代码签名（一种免费的替代方案，无需付费的 Apple 开发者证书）。首次启动应用时：\n\n1. 下载适用于您架构的 `.dmg` 文件（Apple Silicon 使用 aarch64）\n2. 打开 DMG 文件并将应用拖至“应用程序”文件夹\n3. **右键单击**（或按住 Control 键并单击）应用程序，选择“打开”\n4. 在弹出的安全性对话框中再次点击“打开”\n5. 此后应用即可正常启动（后续启动可直接双击）\n\n之所以需要此操作，是因为该应用未使用付费的 Apple 开发者账户进行公证。不过，该应用是安全的——这在通过 App Store 之外分发的开源 macOS 应用中很常见。\n\n## 功能\n\n- 自动从本地或远程 Ollama 服务器获取模型；\n- 遍历多种不同模型、提示和参数以生成推理结果；\n- 同时对多个模型测试不同提示的 A\u002FB 效果；\n- 允许对每组参数组合进行多次迭代；\n- 支持[有限并发](https:\u002F\u002Fdezoito.github.io\u002F2024\u002F03\u002F21\u002Freact-limited-concurrency.html) **或**同步推理调用（以避免过度占用服务器）；\n- 可选输出推理参数及响应元数据（推理时间、token 数及 token\u002F秒）；\n- 支持重新获取单次推理结果；\n- 模型选择可按名称筛选；\n- 实验列表可导出为 JSON 格式；\n- 实验可在可读视图中查看；\n- 可重新运行过往实验，克隆或修改过去使用的参数；\n- 可配置推理超时；\n- 设置中可自定义默认参数与系统提示；\n- 完整的功能性提示数据库，附带示例；\n- 输入时可通过键入“\u002F”实现提示的自动补全。\n\n## 网格搜索（或类似概念...）\n\n严格来说，“网格搜索”是指遍历一系列不同的模型超参数以优化模型性能，但通常指的是诸如 `batch_size`、`learning_rate` 或 `number_of_epochs` 等更常用于训练的参数。\n\n不过，此处的概念与此类似：\n\n让我们定义一组模型、一个提示以及若干参数组合：\n\n![网格参数动画](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdezoito_ollama-grid-search_readme_deb35749a427.gif)\n\n每个参数**值**都会针对选定的每种模型提交一次提示，从而生成一组响应。\n\n## A\u002FB 测试\n\n同样地，您可以通过选择不同模型来执行 A\u002FB 测试，并比较同一提示\u002F参数组合的结果，或者在相似配置下测试不同提示：\n\n![A\u002FB 测试动画](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdezoito_ollama-grid-search_readme_cb21de788495.gif)\n\n\u003Csmall>比较同一模型的不同提示结果\u003C\u002Fsmall>\n\n## 提示档案\n\n您可以保存并管理自己的提示（我们希望使提示与 [Open WebUI](https:\u002F\u002Fgithub.com\u002Fopen-webui\u002Fopen-webui) 兼容）。\n\n![提示档案界面](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdezoito_ollama-grid-search_readme_0ffe38c6b4be.png)\n\n您还可以通过键入“\u002F”实现提示的**自动补全**（灵感来自 Open WebUI）：\n\n![自动补全动画](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdezoito_ollama-grid-search_readme_827153289fd3.gif)\n\n## 实验日志\n\n您可以列出、查看或下载您的实验：\n\n![实验列表界面](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdezoito_ollama-grid-search_readme_7137902746a3.png)\n\n## 未来功能\n\n- 结果评分与按分数筛选\n- 导入、导出及共享提示列表与实验文件。\n\n## 贡献\n\n- 对于明显的 bug 和拼写错误，请直接提交 PR。\n- 如果您想提出新功能、更改现有功能，或提出任何更复杂的建议，请先开启议题进行讨论，**再**着手编写 PR。\n\n## 开发\n\n[开发说明](.\u002Fdocs\u002FDEVELOPMENT.md) 提供了设置指南、序列图和工作流图表，有助于您更好地理解项目并快速上手。\n\n## 引用\n\n以下著作与论文引用了本仓库：\n\nInouye, D & Lindo, L & Lee, R & Allen, E；计算机科学与工程专业毕业论文：**LoRA 超参数的自动调优应用**\n圣塔克拉拉大学，2024年\n\u003Chttps:\u002F\u002Fscholarcommons.scu.edu\u002Fcgi\u002Fviewcontent.cgi?article=1271&context=cseng_senior>\n\n## 谢谢！\n\n特别感谢 [@FabianLars](https:\u002F\u002Fgithub.com\u002FFabianLars)、[@peperroni21](https:\u002F\u002Fgithub.com\u002Fpepperoni21) 和 [@TomReidNZ](https:\u002F\u002Fgithub.com\u002FTomReidNZ)。","# ollama-grid-search 快速上手指南\n\n## 环境准备\n\n### 系统要求\n- macOS（支持 Apple Silicon，如 M1\u002FM2\u002FM3\u002FM4）\n- Windows\n- Linux\n\n### 前置依赖\n- 已安装 [Ollama](https:\u002F\u002Fwww.ollama.ai)，并确保其在本地或远程服务器上运行。\n\n---\n\n## 安装步骤\n\n### macOS 安装（Apple Silicon）\n\n1. 下载适用于你架构的 `.dmg` 文件（Apple Silicon 使用 `aarch64`）。\n2. 双击打开 DMG 文件，将应用拖拽到“Applications”文件夹中。\n3. **右键点击**（或使用 `Control+click`）应用程序，选择 **\"Open\"**。\n4. 在出现的安全提示中点击 **\"Open\"**。\n5. 应用即可正常启动（后续可直接双击打开）。\n\n> 注意：由于未使用付费 Apple 开发者证书进行公证，首次启动时需要手动允许。该应用是安全的，这是开源 macOS 应用的标准做法。\n\n### 其他系统安装\n\n前往项目 [发布页面](https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Freleases) 下载对应系统的安装包：\n\n- [macOS](https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Freleases\u002Flatest)\n- [Windows](https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Freleases\u002Flatest)\n- [Linux](https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Freleases\u002Flatest)\n- [所有版本](https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Freleases)\n\n---\n\n## 基本使用\n\n### 最简单的使用示例\n\n1. 打开 `ollama-grid-search` 应用。\n2. 在主界面中，选择你要测试的模型（例如 `llama3`, `mistral`, `phi3`）。\n3. 输入一个测试提示（prompt），例如：\n   ```\n   请解释什么是量子计算？\n   ```\n4. 设置参数组合（如温度、最大 token 数等）。\n5. 点击 **开始实验**，工具将自动为每个模型生成响应，并展示结果对比。\n\n> 提示：你可以通过输入 `\u002F` 来快速调用预定义的提示模板。\n\n---\n\n通过以上步骤，你可以快速上手使用 `ollama-grid-search` 进行多模型、多提示的 A\u002FB 测试和参数搜索实验。","某科技公司的内容团队正在为一款智能客服系统选择最适合的大型语言模型（LLM），以提升用户问题的准确理解和响应质量。他们需要在多个模型和不同提示词之间进行比较，以找到最佳组合。\n\n### 没有 ollama-grid-search 时\n\n- 需要手动逐个测试不同的模型和提示词组合，耗时且容易出错；\n- 无法同时对比多个模型在同一任务下的表现，难以直观发现差异；\n- 缺乏对推理参数（如温度、最大token数）的系统性调整与记录；\n- 每次实验结果需手动整理和分析，效率低下；\n- 无法保存和复用之前的实验配置，重复工作量大。\n\n### 使用 ollama-grid-search 后\n\n- 可以一键定义多个模型、提示词及参数组合，自动并行运行实验，节省大量时间；\n- 提供可视化界面，可同时对比不同模型在相同任务下的输出结果和性能指标；\n- 支持对推理参数进行系统化调整，并记录每次实验的详细信息（如推理时间、token数量等）；\n- 实验结果可导出为JSON格式，便于后续分析和归档，也可直接在界面上重新运行或修改；\n- 提供“提示库”功能，支持快速调用和自定义提示模板，提升工作效率。\n\nollama-grid-search 让模型选择和优化过程更加高效、直观和可重复。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdezoito_ollama-grid-search_4c07171c.png","dezoito","Dezoito","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Fdezoito_8169d706.png","Building airgapped ML systems\r\n| MSc CS\u002FAI student ",null,"https:\u002F\u002Fdezoito.github.io","https:\u002F\u002Fgithub.com\u002Fdezoito",[83,87,91,95,99],{"name":84,"color":85,"percentage":86},"TypeScript","#3178c6",84.3,{"name":88,"color":89,"percentage":90},"Rust","#dea584",13.4,{"name":92,"color":93,"percentage":94},"JavaScript","#f1e05a",1.1,{"name":96,"color":97,"percentage":98},"CSS","#663399",0.8,{"name":100,"color":101,"percentage":102},"HTML","#e34c26",0.4,921,55,"2026-03-30T13:06:12","MIT","Linux, macOS, Windows","未说明",{"notes":110,"python":108,"dependencies":111},"该工具需要安装 Ollama 并运行在本地或远程服务器上。对于 macOS 用户，首次运行时需通过右键点击应用并选择“打开”以绕过安全限制。",[],[14,13,26,15],[114,115,116,117,118,119,120,121],"ai","gridsearch","llm","ollama","rust","ab-testing","testing-tools","grid-search","2026-03-27T02:49:30.150509","2026-04-06T07:16:06.772439",[125,130,135,140,145,150],{"id":126,"question_zh":127,"answer_zh":128,"source_url":129},5598,"连接到远程 Ollama 服务器时出现连接问题","如果使用默认端口（如 https:\u002F\u002Follama.my.internet.network）连接远程 Ollama 服务器时遇到问题，可以尝试在 URL 中显式指定端口号。版本 0.5.3 已修复此问题，若未指定端口，将自动使用 443（HTTPS）或 80（HTTP）。","https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Fissues\u002F39",{"id":131,"question_zh":132,"answer_zh":133,"source_url":134},5599,"如何解决 macOS 上的“只读文件系统”错误？","在 macOS 上运行时，如果遇到 `Read-only file system (os error 30)` 错误，请尝试用逗号分隔参数值而不是空格。例如：`--param1 value1,value2`。这可以避免因输入验证导致的问题。","https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Fissues\u002F4",{"id":136,"question_zh":137,"answer_zh":138,"source_url":139},5600,"如何解决 macOS 应用程序被阻止打开的问题？","如果 macOS 报错说应用程序损坏且无法打开，请尝试完全卸载旧版本后重新安装最新版本。或者尝试使用 x64 安装程序并启用 Rosetta 2 进行兼容性运行。","https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Fissues\u002F16",{"id":141,"question_zh":142,"answer_zh":143,"source_url":144},5601,"Windows 上出现 'failed to create context with model' 错误怎么办？","该错误可能与 Ollama 在 Windows 上的兼容性有关。请检查 Ollama 的官方仓库中是否有相关修复（如 #3774 和 #3850），并确保模型路径正确无误。","https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Fissues\u002F17",{"id":146,"question_zh":147,"answer_zh":148,"source_url":149},5602,"为什么会出现 'chat template not supported' 错误？","Ollama 当前不支持某些自定义 chat template，会回退到默认的 chatml 模板。建议使用 `\u002Fchat` 端点进行交互，而非 `\u002Fgenerate` 端点，以获得更好的兼容性。","https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Fissues\u002F7",{"id":151,"question_zh":152,"answer_zh":153,"source_url":154},5603,"如何解决 macOS 上下载的 dmg 文件被 Apple 验证失败的问题？","尝试更新到最新版本（如 0.9.2），该版本已修复此问题。如果仍然存在问题，可以尝试从其他来源下载或联系开发者获取帮助。","https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search\u002Fissues\u002F70",[156,161,166,171,176,181,186,190,195,200,205,210,215,220,225,230,235,240,245,250],{"id":157,"version":158,"summary_zh":159,"released_at":160},105217,"v0.9.2","## [Version 0.9.2] - 2025-11-20\r\n\r\n### Changed\r\n\r\n- Bumps ollama-rs to versoin 0.3.3\r\n- Attempts to fix damaged .dmg files generated during release github action\r\n\r\n**macOS Apple Silicon Users:** After downloading, right-click the app and select \"Open\" on first launch to bypass Gatekeeper. This is required because the app uses ad-hoc code signing (no paid Apple Developer certificate). Subsequent launches work normally.","2025-11-20T22:59:29",{"id":162,"version":163,"summary_zh":164,"released_at":165},105218,"v0.9.1","## [Version 0.9.1] - 2025-04-17\r\n\r\n### Changed\r\n\r\n- Bumps ollama-rs to versoin 0.3.0","2025-04-17T18:44:47",{"id":167,"version":168,"summary_zh":169,"released_at":170},105219,"v0.9.0","## [Version 0.9.0] - 2024-12-20\r\n\r\n### Added\r\n\r\n- Support for navigating over [variable] placeholders in prompts, to paste input.\r\n- Tooltips in upper menu corner buttons.\r\n\r\n### Changed\r\n\r\n- Renamed components that had stupid names.\r\n- Removed MacOS 12 from the supported OSes when building releases (blame GH Actions).\r\n- Experimentally added support for MacOS 13.","2024-12-20T14:11:50",{"id":172,"version":173,"summary_zh":174,"released_at":175},105220,"v0.8.0","Notice: MacOS 12 is not supported by GitHub actions any longer, therefore we are only releasing versions for MacOS-latest!\r\n\r\n## [Version 0.8.0] - 2024-12-08\r\n\r\n### Changed\r\n\r\n- Experiments are stored in a database. File system is not used anymore\r\n- Minor UI improvements in the Experiment selecion UI","2024-12-08T13:50:05",{"id":177,"version":178,"summary_zh":179,"released_at":180},105221,"v0.7.0","## [Version 0.7.0] - 2024-11-24\r\n\r\n### Added\r\n\r\n- Fully functional prompt management UI\r\n- Prompt inputs can trigger autocomplete by starting with \"\u002F\"\r\n- Added SQLite integration for prompt management and other developments\r\n\r\n### Changed\r\n\r\n- Several small UI improvements (mostly using ScrollAreas instead of overflows\r\n- Improved validation rules on experiment form","2024-11-25T00:24:16",{"id":182,"version":183,"summary_zh":184,"released_at":185},105222,"v0.7.0-rc2","See the assets to download this version and install.","2024-11-25T00:07:19",{"id":187,"version":188,"summary_zh":184,"released_at":189},105223,"v0.7.0-rc1","2024-11-24T23:40:49",{"id":191,"version":192,"summary_zh":193,"released_at":194},105224,"v0.6.2","## [Version 0.6.2] - 2024-10-29\r\n\r\n### Fixed\r\n\r\n- The \"refetch\" button must be shown when there was an error in the inference call.","2024-10-29T22:27:22",{"id":196,"version":197,"summary_zh":198,"released_at":199},105225,"v0.6.1","## [Version 0.6.1] - 2024-10-28\r\n\r\n### Changed\r\n\r\n- When removing all experiment logs, only JSON files should be deleted.\r\n- Add colors to prompt and system_prompt when displaying inference params in results.\r\n- Border colors are used on the side of a result to group outputs from the same model.","2024-10-29T00:49:53",{"id":201,"version":202,"summary_zh":203,"released_at":204},105226,"v0.6.0","## [Version 0.6.0] - 2024-10-20\r\n\r\n### Added\r\n\r\n- Added UI controls to re-run past experiments.\r\n- Added controls to remove experiment files from the UI.\r\n- Added button to copy an inference text to the clipboard.\r\n\r\n### Changed\r\n\r\n- Moved \"reload\" icon to improve layout.\r\n- Improved experiment inspection UI readability.\r\n- Streamlined State management.\r\n\r\n### Fixes\r\n\r\n- Fix HMR not working on MacOS (in development, of course).\r\n","2024-10-20T13:35:16",{"id":206,"version":207,"summary_zh":208,"released_at":209},105227,"v0.5.3","## [Version 0.5.3] - 2024-09-16\r\n\r\n### Fixes\r\n\r\n- Handles Ollama servers using default ports (80 or 443)","2024-09-16T21:46:02",{"id":211,"version":212,"summary_zh":213,"released_at":214},105228,"v0.5.2","## [Version 0.5.2] - 2024-09-15\r\n\r\n### Added\r\n\r\n - Adds custom application icon.\r\n\r\n### Fixes\r\n\r\n - Handles Ollama version info not being correctly returned by the server.","2024-09-15T14:48:10",{"id":216,"version":217,"summary_zh":218,"released_at":219},105229,"v0.5.1","## [Version 0.5.1] - 2024-07-10\r\n\r\n### Added\r\n\r\n- Added Clippy checks when saving Rust code.\r\n- Corrected existing Rust code to pass Clippy checks.\r\n- Improved UI for component that displays inference parameters with collapsible prompts.\r\n\r\n### Changed\r\n\r\n- Fixes generation responses not returning metadata (like `eval_duration`, `total_duration`, `eval_count`).\r\n- Added Rust CI checks.\r\n- Fixed padding in \"Expand\u002FHide\" buttons for params and metadate.\r\n- `keep_alive` parameter for generation is set to Ollama's default (instead of `indefinitely`).","2024-07-11T00:16:27",{"id":221,"version":222,"summary_zh":223,"released_at":224},105230,"v0.5.0","## [Version 0.5.0] - 2024-05-10\r\n\r\n### Added\r\n\r\n- Allows multiple prompts when running experiments (courtesy of @calebsheridan).\r\n- Allows multiple concurrent inference calls, matching Ollama's support for concurrency.\r\n- Allows user to hide model names on the results pane, to avoid bias in evaluations.\r\n- Added `concurrent_inferences` input to settings.\r\n- Added `hide_model_names` input to settings.\r\n\r\n### Changed\r\n\r\n- Experiment inspection view was updated to show which prompt was used for each iteration, preserving their line breaks (courtesy of @calebsheridan).\r\n- Minor UI tweaks.","2024-05-10T13:54:40",{"id":226,"version":227,"summary_zh":228,"released_at":229},105231,"v0.4.3","## [Version 0.4.3] - 2024-05-01\r\n\r\n### Fixes\r\n\r\n- Bug in new installations that had no previous configs for \"system prompt\"","2024-05-01T19:32:01",{"id":231,"version":232,"summary_zh":233,"released_at":234},105232,"v0.4.2","## [Version 0.4.2] - 2024-05-01\r\n\r\n### Added\r\n\r\n- Adds basic self-signed code signing to macOS app\u002FDMG.\r\n\r\n### Changed\r\n\r\n- Bumps Linux build runner version to current stable.\r\n- Builds for both ARM and x86 macOS separately.\r\n- Bumps Tauri package versions to current stable.\r\n\r\nAll contributions from @sammcj!","2024-05-01T17:18:19",{"id":236,"version":237,"summary_zh":238,"released_at":239},105233,"v0.4.1","## [Version 0.4.1] - 2024-04-27\r\n\r\n### Changed\r\n\r\n- Github action now builds releases for INTEL and AARCH based Macs.\r\n- Bumped `ollama-rs` to version 0.1.9.\r\n- Improved readability for past experiments.","2024-04-27T19:03:13",{"id":241,"version":242,"summary_zh":243,"released_at":244},105234,"v0.4.0","## [Version 0.4.0] - 2024-04-26\r\n\r\n### Added\r\n\r\n- Added system prompt to main form\r\n- Added versioning to the JSON log files, starting at this release's version.\r\n\r\n### Changed\r\n\r\n- Many UI improvements (Thanks to @calebsheridan)\r\n- Fixes issue with date of the experiment being updated on re-renders\r\n- Presents experiment inspection output in human readable format instead of JSON\r\n- Fixes bug where the experiment's start date would be updated on component's re-render.","2024-04-26T19:49:31",{"id":246,"version":247,"summary_zh":248,"released_at":249},105235,"v0.3.0","## [Version 0.3.0] - 2024-04-12\r\n\r\n### Added\r\n\r\n- Allow multiple generations for each combination of parameters.\r\n- Improved metrics are displayed after inference results\r\n- Allows inspecting past experiments without the need to download the corresponding file.\r\n- Displays current version number to the settings interface.\r\n- Displays the version of Ollama running on the selected server","2024-04-12T16:25:05",{"id":251,"version":252,"summary_zh":184,"released_at":253},105236,"v0.2.1","2024-03-24T13:47:47"]