[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-probabl-ai--skore":3,"tool-probabl-ai--skore":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":80,"owner_website":81,"owner_url":82,"languages":83,"stars":110,"forks":111,"last_commit_at":112,"license":113,"difficulty_score":114,"env_os":115,"env_gpu":116,"env_ram":116,"env_deps":117,"category_tags":122,"github_topics":123,"view_count":10,"oss_zip_url":79,"oss_zip_packed_at":79,"status":16,"created_at":131,"updated_at":132,"faqs":133,"releases":164},710,"probabl-ai\u002Fskore","skore","Track your Data Science. Skore's open-source Python library accelerates ML model development with automated evaluation reports, smart methodological guidance, and comprehensive cross-validation analysis.","skore 是一款开源 Python 库，专注于优化机器学习模型的开发生命周期。它通过自动生成评估报告、提供智能的方法论指导以及全面的交叉验证分析，帮助数据科学家高效追踪和管理数据科学项目。\n\n在使用 pandas 或 scikit-learn 等通用工具时，开发者常需花费大量时间编写样板代码、查阅文档并维护项目结构。skore 充当了“指挥者”的角色，将非结构化的开发过程转化为结构化的实验成果。其核心亮点在于仅需一行代码即可获取综合模型评估洞察，内置的方法论警告还能有效避免常见陷阱，确保实验质量。\n\nskore 特别适合机器学习工程师和数据研究人员使用。除了核心的 Python 库（Skore Lib），它还集成了协作平台（Skore Hub），方便团队分享和比较实验结果。如果你希望提升实验的可复现性、减少重复劳动并保持项目清晰有序，skore 是一个非常实用的选择。","\u003Cdiv align=\"center\">\n\n  ![license](https:\u002F\u002Fimg.shields.io\u002Fpypi\u002Fl\u002Fskore)\n  ![python](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fpython-3.10%20%7C%203.11%20%7C%203.12%20%7C%203.13-blue?style=flat&logo=python)\n  [![downloads](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fprobabl-ai_skore_readme_e553369f4034.png)](https:\u002F\u002Fpepy.tech\u002Fprojects\u002Fskore)\n  [![pypi](https:\u002F\u002Fimg.shields.io\u002Fpypi\u002Fv\u002Fskore)](https:\u002F\u002Fpypi.org\u002Fproject\u002Fskore\u002F)\n  [![Discord](https:\u002F\u002Fimg.shields.io\u002Fdiscord\u002F1275821367324840119?label=Discord)](https:\u002F\u002Fdiscord.probabl.ai\u002F)\n\n\u003C\u002Fdiv>\n\n\u003Cdiv align=\"center\">\n\n  \u003Cpicture>\n    \u003Csource srcset=\"https:\u002F\u002Fmedia.githubusercontent.com\u002Fmedia\u002Fprobabl-ai\u002Fskore\u002Fmain\u002Fsphinx\u002F_static\u002Fimages\u002FLogo_Skore_Dark@2x.svg\" media=\"(prefers-color-scheme: dark)\">\n    \u003Cimg width=\"200\" src=\"https:\u002F\u002Fmedia.githubusercontent.com\u002Fmedia\u002Fprobabl-ai\u002Fskore\u002Fmain\u002Fsphinx\u002F_static\u002Fimages\u002FLogo_Skore_Light@2x.svg\" alt=\"skore logo\">\n  \u003C\u002Fpicture>\n  \u003Ch3>Track Your Data Science\u003C\u002Fh3>\n\nElevate ML Development with Built-in Recommended Practices \\\n[Documentation](https:\u002F\u002Fdocs.skore.probabl.ai) — [Community](https:\u002F\u002Fdiscord.probabl.ai) — [YouTube](https:\u002F\u002Fyoutube.com\u002Fplaylist?list=PLSIzlWDI17bTpixfFkooxLpbz4DNQcam3) — [Skore Hub](https:\u002F\u002Fprobabl.ai\u002Fskore)\n\n\u003C\u002Fdiv>\n\n\u003Cbr \u002F>\n\n## 🎯 Why Skore?\n\nWhen it comes to data science, you have excellent tools at your disposal: `pandas` and `polars` for data exploration, `skrub` for stateful transformations, and `scikit-learn` for model training and evaluation. These libraries are designed to be generic and accommodate a wide range of use cases.\n\n**But here's the challenge**: Your experience is key to choosing the right building blocks and methodologies. You often spend significant time navigating documentation, writing boilerplate code for common evaluations, and struggling to maintain clear project structure.\n\n**Skore is the conductor** that transforms your data science pipeline into structured, meaningful artifacts. It reduces the time you spend on documentation navigation, eliminates boilerplate code, and guides you toward the right methodological information to answer your questions.\n\n### What Skore does for you:\n\n- **Structures your experiments**: Automatically generates the insights that matter for your use case\n- **Reduces boilerplate**: One line of code gives you comprehensive model evaluation\n- **Guides your decisions**: Built-in methodological warnings help you avoid common pitfalls\n- **Maintains clarity**: Structured project organization makes your work easier to understand and maintain\n\n⭐ Support us with a star and spread the word - it means a lot! ⭐\n\n## 🧩 What is Skore?\n\nThe core mission of **Skore** is to turn uneven ML development into structured, effective decision-making. It consists of two complementary components:\n- **Skore Lib**: the open-source Python library (described here!) that provides the structured artifacts and methodological guidance for your data science experiments.\n- **Skore Hub**: the collaborative platform where teams can share, compare, and build upon each other's structured experiments. Learn more on our [product page](https:\u002F\u002Fprobabl.ai\u002Fskore).\n\n## ⚡️ Quick start\n\n### Installation\n\n#### With pip\n\nWe recommend using a [virtual environment (venv)](https:\u002F\u002Fdocs.python.org\u002F3\u002Ftutorial\u002Fvenv.html). You need `python>=3.10`.\n\nThen, you can install skore by using `pip`:\n```bash\n# If you plan to use Skore locally\npip install -U skore\n# If you wish to interact with Skore Hub as well\npip install -U skore[hub]\n# If you wish to log projects to MLflow\npip install -U skore[mlflow]\n```\n\n#### With conda\n\nskore is available in `conda-forge` both for local and hub use:\n\n```bash\nconda install conda-forge::skore\n```\n\nYou can find information on the latest version [here](https:\u002F\u002Fanaconda.org\u002Fconda-forge\u002Fskore).\n\n### Get structured insights from your ML pipeline\n\nEvaluate your model and get comprehensive insights in one line:\n\n```python\nfrom sklearn.datasets import make_classification\nfrom sklearn.linear_model import LogisticRegression\nfrom skore import CrossValidationReport\n\nX, y = make_classification(n_classes=2, n_samples=100_000, n_informative=4)\nclf = LogisticRegression()\n\n# Get structured insights that matter for your use case\ncv_report = CrossValidationReport(clf, X, y)\n\n# See what insights are available\ncv_report.help()\n\n# Example: Access the metrics summary\nmetrics_summary = cv_report.metrics.summarize().frame()\n\n# Example: Get the ROC curve\nroc_plot = cv_report.metrics.roc()\nroc_plot.plot()\n```\n\nLearn more in our [documentation](https:\u002F\u002Fdocs.skore.probabl.ai).\n\n## 🛠️ Contributing\n\nJoin our mission to promote open-source and make machine learning development more robust and effective. If you'd like to contribute, please check the contributing guidelines [here](https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fblob\u002Fmain\u002FCONTRIBUTING.rst).\n\n## 👋 Feedback & Community\n\n-   Join our [Discord](https:\u002F\u002Fdiscord.probabl.ai\u002F) to share ideas or get support.\n-   Request a feature or report a bug via [GitHub Issues](https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fissues).\n\n## Support\n\nSkore is tested on Linux and Windows, for at most 4 versions of Python, and at most 4 versions of scikit-learn:\n- Python 3.10\n  - scikit-learn 1.5\n  - scikit-learn 1.7\n- Python 3.11\n  - scikit-learn 1.5\n  - scikit-learn 1.8\n- Python 3.12\n  - scikit-learn 1.5\n  - scikit-learn 1.8\n- Python 3.13\n  - scikit-learn 1.5\n  - scikit-learn 1.6\n  - scikit-learn 1.7\n  - scikit-learn 1.8\n\n---\n\nBrought to you by\n\n\u003Ca href=\"https:\u002F\u002Fprobabl.ai\u002Fskore\" target=\"_blank\">\n    \u003Cpicture>\n        \u003Csource srcset=\"https:\u002F\u002Fmedia.githubusercontent.com\u002Fmedia\u002Fprobabl-ai\u002Fskore\u002Fmain\u002Fsphinx\u002F_static\u002Fimages\u002FProbabl-logo-orange.png\" media=\"(prefers-color-scheme: dark)\">\n        \u003Cimg width=\"120\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fprobabl-ai_skore_readme_32871deceb87.png\" alt=\"Probabl logo\">\n    \u003C\u002Fpicture>\n\u003C\u002Fa>\n","\u003Cdiv align=\"center\">\n\n  ![license](https:\u002F\u002Fimg.shields.io\u002Fpypi\u002Fl\u002Fskore)\n  ![python](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fpython-3.10%20%7C%203.11%20%7C%203.12%20%7C%203.13-blue?style=flat&logo=python)\n  [![downloads](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fprobabl-ai_skore_readme_e553369f4034.png)](https:\u002F\u002Fpepy.tech\u002Fprojects\u002Fskore)\n  [![pypi](https:\u002F\u002Fimg.shields.io\u002Fpypi\u002Fv\u002Fskore)](https:\u002F\u002Fpypi.org\u002Fproject\u002Fskore\u002F)\n  [![Discord](https:\u002F\u002Fimg.shields.io\u002Fdiscord\u002F1275821367324840119?label=Discord)](https:\u002F\u002Fdiscord.probabl.ai\u002F)\n\n\u003C\u002Fdiv>\n\n\u003Cdiv align=\"center\">\n\n  \u003Cpicture>\n    \u003Csource srcset=\"https:\u002F\u002Fmedia.githubusercontent.com\u002Fmedia\u002Fprobabl-ai\u002Fskore\u002Fmain\u002Fsphinx\u002F_static\u002Fimages\u002FLogo_Skore_Dark@2x.svg\" media=\"(prefers-color-scheme: dark)\">\n    \u003Cimg width=\"200\" src=\"https:\u002F\u002Fmedia.githubusercontent.com\u002Fmedia\u002Fprobabl-ai\u002Fskore\u002Fmain\u002Fsphinx\u002F_static\u002Fimages\u002FLogo_Skore_Light@2x.svg\" alt=\"skore logo\">\n  \u003C\u002Fpicture>\n  \u003Ch3>追踪你的数据科学 (Data Science)\u003C\u002Fh3>\n\n通过内置最佳实践 (Best Practices) 提升机器学习 (Machine Learning) 开发 \\\n[文档](https:\u002F\u002Fdocs.skore.probabl.ai) — [社区](https:\u002F\u002Fdiscord.probabl.ai) — [YouTube](https:\u002F\u002Fyoutube.com\u002Fplaylist?list=PLSIzlWDI17bTpixfFkooxLpbz4DNQcam3) — [Skore Hub](https:\u002F\u002Fprobabl.ai\u002Fskore)\n\n\u003C\u002Fdiv>\n\n\u003Cbr \u002F>\n\n## 🎯 为什么选择 Skore？\n\n在数据科学 (Data Science) 领域，你拥有出色的工具可供使用：用于数据探索的 `pandas` 和 `polars`，用于有状态转换的 `skrub`，以及用于模型训练和评估的 `scikit-learn`。这些库设计为通用的，以适应广泛的使用场景。\n\n**但挑战在于：** 你的经验对于选择合适的构建模块和方法论至关重要。你经常需要花费大量时间浏览文档，为常见评估编写样板代码 (Boilerplate Code)，并努力维持清晰的项目结构。\n\n**Skore 就是那位指挥家**，它将你的数据科学流水线 (Pipeline) 转化为结构化的、有意义的工件 (Artifacts)。它减少了你在文档导航上花费的时间，消除了样板代码，并引导你获取正确的方法论信息以回答你的问题。\n\n### Skore 为你做的事情：\n\n- **结构化你的实验**：自动生成对你使用场景至关重要的洞察\n- **减少样板代码**：一行代码即可实现全面的模型评估\n- **指导你的决策**：内置的方法论警告帮助你避免常见陷阱\n- **保持清晰性**：结构化的项目组织使你的工作更易于理解和维护\n\n⭐ 请用 Star 支持我们并传播消息 - 这意义重大！⭐\n\n## 🧩 什么是 Skore？\n\n**Skore** 的核心使命是将非结构化的机器学习 (Machine Learning) 开发转化为结构化的、有效的决策制定。它由两个互补的组件组成：\n- **Skore Lib**：开源 Python 库（此处描述！），为你的数据科学实验提供结构化工件和方法论指导。\n- **Skore Hub**：协作平台，团队可以在其中共享、比较并基于彼此的结构化实验进行构建。更多详情请访问我们的 [产品页面](https:\u002F\u002Fprobabl.ai\u002Fskore)。\n\n## ⚡️ 快速开始\n\n### 安装\n\n#### 使用 pip\n\n我们建议使用 [虚拟环境 (venv)](https:\u002F\u002Fdocs.python.org\u002F3\u002Ftutorial\u002Fvenv.html)。你需要 `python>=3.10`。\n\n然后，你可以使用 `pip` 安装 skore：\n```bash\n# 如果你计划本地使用 Skore\npip install -U skore\n# 如果你也希望与 Skore Hub 交互\npip install -U skore[hub]\n# 如果你希望将项目记录到 MLflow\npip install -U skore[mlflow]\n```\n\n#### 使用 conda\n\nskore 在 `conda-forge` 中可用，适用于本地和 Hub 使用：\n\n```bash\nconda install conda-forge::skore\n```\n\n你可以在 [这里](https:\u002F\u002Fanaconda.org\u002Fconda-forge\u002Fskore) 找到关于最新版本的信息。\n\n### 从你的机器学习流水线 (ML Pipeline) 获取结构化洞察\n\n在一行代码内评估你的模型并获得全面洞察：\n\n```python\nfrom sklearn.datasets import make_classification\nfrom sklearn.linear_model import LogisticRegression\nfrom skore import CrossValidationReport\n\nX, y = make_classification(n_classes=2, n_samples=100_000, n_informative=4)\nclf = LogisticRegression()\n\n# 获取对你使用场景至关重要的结构化洞察\ncv_report = CrossValidationReport(clf, X, y)\n\n# 查看可用的洞察\ncv_report.help()\n\n# 示例：访问指标摘要\nmetrics_summary = cv_report.metrics.summarize().frame()\n\n# 示例：获取 ROC 曲线\nroc_plot = cv_report.metrics.roc()\nroc_plot.plot()\n```\n\n在我们的 [文档](https:\u002F\u002Fdocs.skore.probabl.ai) 中了解更多。\n\n## 🛠️ 贡献\n\n加入我们的使命，推广开源并使机器学习开发更加稳健和高效。如果你想贡献，请查看贡献指南 [这里](https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fblob\u002Fmain\u002FCONTRIBUTING.rst)。\n\n## 👋 反馈与社区\n\n-   加入我们的 [Discord](https:\u002F\u002Fdiscord.probabl.ai\u002F) 分享想法或获取支持。\n-   通过 [GitHub Issues](https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fissues) 请求功能或报告错误。\n\n## 支持\n\nSkore 在 Linux 和 Windows 上进行了测试，针对最多 4 个版本的 Python 和最多 4 个版本的 scikit-learn：\n- Python 3.10\n  - scikit-learn 1.5\n  - scikit-learn 1.7\n- Python 3.11\n  - scikit-learn 1.5\n  - scikit-learn 1.8\n- Python 3.12\n  - scikit-learn 1.5\n  - scikit-learn 1.8\n- Python 3.13\n  - scikit-learn 1.5\n  - scikit-learn 1.6\n  - scikit-learn 1.7\n  - scikit-learn 1.8\n\n---\n\n由以下人员提供\n\n\u003Ca href=\"https:\u002F\u002Fprobabl.ai\u002Fskore\" target=\"_blank\">\n    \u003Cpicture>\n        \u003Csource srcset=\"https:\u002F\u002Fmedia.githubusercontent.com\u002Fmedia\u002Fprobabl-ai\u002Fskore\u002Fmain\u002Fsphinx\u002F_static\u002Fimages\u002FProbabl-logo-orange.png\" media=\"(prefers-color-scheme: dark)\">\n        \u003Cimg width=\"120\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fprobabl-ai_skore_readme_32871deceb87.png\" alt=\"Probabl logo\">\n    \u003C\u002Fpicture>\n\u003C\u002Fa>","# skore 快速上手指南\n\n**skore** 是一款用于跟踪数据科学项目的工具，旨在通过内置的最佳实践提升机器学习开发效率。它能将实验转化为结构化的成果，减少样板代码，并指导决策。\n\n## 1. 环境准备\n\n在开始之前，请确保您的开发环境满足以下要求：\n\n- **操作系统**：Linux 或 Windows\n- **Python 版本**：>= 3.10（支持 3.10, 3.11, 3.12, 3.13）\n- **依赖管理**：建议使用虚拟环境（venv）隔离依赖\n\n## 2. 安装步骤\n\n### 使用 pip 安装\n\n推荐使用虚拟环境进行安装。根据您的需求选择相应的命令：\n\n```bash\n# 仅本地使用\npip install -U skore\n\n# 如需与 Skore Hub 交互\npip install -U skore[hub]\n\n# 如需将项目记录到 MLflow\npip install -U skore[mlflow]\n```\n\n### 使用 conda 安装\n\nskore 也提供于 `conda-forge` 通道中：\n\n```bash\nconda install conda-forge::skore\n```\n\n## 3. 基本使用\n\nskore 的核心功能是通过一行代码获取模型的综合评估洞察。以下是基于 `CrossValidationReport` 的最小化示例：\n\n```python\nfrom sklearn.datasets import make_classification\nfrom sklearn.linear_model import LogisticRegression\nfrom skore import CrossValidationReport\n\nX, y = make_classification(n_classes=2, n_samples=100_000, n_informative=4)\nclf = LogisticRegression()\n\n# 获取针对您用例的结构化洞察\ncv_report = CrossValidationReport(clf, X, y)\n\n# 查看可用的洞察内容\ncv_report.help()\n\n# 示例：访问指标摘要\nmetrics_summary = cv_report.metrics.summarize().frame()\n\n# 示例：获取 ROC 曲线\nroc_plot = cv_report.metrics.roc()\nroc_plot.plot()\n```\n\n更多详细信息请参阅 [官方文档](https:\u002F\u002Fdocs.skore.probabl.ai)。","某金融科技公司数据工程师正在开发信贷违约预测模型，需要在短时间内完成多轮实验并验证模型稳定性。\n\n### 没有 skore 时\n- 需要手动编写大量重复的评估代码和可视化脚本，不仅耗时还容易引入人为错误。\n- 跨验证过程缺乏标准化检查，开发者容易忽略数据泄露或过拟合等常见方法论陷阱。\n- 不同版本的实验结果分散在本地文件夹中，难以快速对比历史迭代并追溯最佳模型参数。\n- 向业务方汇报时需要额外花费时间整理文档和图表，沟通成本较高且信息传递不统一。\n\n### 使用 skore 后\n- 仅需一行代码即可生成包含交叉验证分析的完整评估报告，大幅减少了样板代码的编写工作量。\n- 内置的方法论指导会自动提示潜在风险，帮助开发者在编码阶段就避免常见的建模错误。\n- 实验结果被结构化存储，清晰展示每次迭代的关键指标变化，便于随时调取对比。\n- 自动生成的洞察直接可用于汇报，让团队能更专注于核心算法优化而非文档整理工作。\n\nskore 将零散的实验流程转化为结构化的决策依据，显著提升了机器学习开发的效率与可靠性。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fprobabl-ai_skore_135ee3f3.png","probabl-ai",":probabl.","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Fprobabl-ai_abee23eb.png","Own Your Data Science",null,"probabl_ai","https:\u002F\u002Fprobabl.ai\u002F","https:\u002F\u002Fgithub.com\u002Fprobabl-ai",[84,88,92,95,99,103,106],{"name":85,"color":86,"percentage":87},"Python","#3572A5",96.6,{"name":89,"color":90,"percentage":91},"CSS","#663399",1.1,{"name":93,"color":94,"percentage":91},"Jinja","#a52a22",{"name":96,"color":97,"percentage":98},"HTML","#e34c26",0.6,{"name":100,"color":101,"percentage":102},"JavaScript","#f1e05a",0.3,{"name":104,"color":105,"percentage":102},"Shell","#89e051",{"name":107,"color":108,"percentage":109},"Makefile","#427819",0.1,624,135,"2026-04-03T15:15:29","MIT",1,"Linux, Windows","未说明",{"notes":118,"python":119,"dependencies":120},"建议使用虚拟环境 (venv)；支持本地运行或连接 Skore Hub；可选安装 [hub] 或 [mlflow] 扩展；官方明确测试平台为 Linux 和 Windows。","3.10+",[121],"scikit-learn>=1.5",[13,54,51,15],[124,125,126,127,128,129,130],"data-science","machine-learning","data-analysis","python","scikit-learn","data-visualization","workflow","2026-03-27T02:49:30.150509","2026-04-06T07:11:52.576450",[134,139,144,149,154,159],{"id":135,"question_zh":136,"answer_zh":137,"source_url":138},2985,"如何创建或加载 skore 项目？","推荐使用新的 skore.open() 函数统一管理项目生命周期。用法示例：my_project = skore.open(\"quick_start\", create=True, overwrite=False)。该函数允许灵活控制是否创建新项目或加载现有项目，无需在 create 和 load 之间切换。","https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fissues\u002F917",{"id":140,"question_zh":141,"answer_zh":142,"source_url":143},2986,"如何在使用 train_test_split 时避免警告？","建议显式传递 X 和 y 数组给函数（如 train_test_split(X, y, ...)），确保系统能准确识别目标变量位置。关于 shuffle 参数，默认行为可能触发警告（假设数据独立同分布），请根据实际数据分布情况谨慎设置，通常建议明确指定 shuffle 状态以避免误报。","https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fissues\u002F492",{"id":145,"question_zh":146,"answer_zh":147,"source_url":148},2987,"如何对最终训练好的模型进行诚实评估？","使用顶层的 skore.evaluate() 函数进行最终测试集评估。示例代码：report = evaluate(final_model, X_test, y_test)，随后调用 report.metrics.summarize()。这比手动创建 EstimatorReport 更简洁且符合开发流程。","https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fissues\u002F2520",{"id":150,"question_zh":151,"answer_zh":152,"source_url":153},2988,"如何在报告中获取模型的执行时间（拟合和评分）？","使用 report.metrics.report_metrics(scoring=\"timing\") 方法。该方法会内部调用 timing 函数并拼接相关信息。你可以通过 scoring 参数控制获取的指标类型，从而查看 fit_time 和 score_time。","https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fissues\u002F1241",{"id":155,"question_zh":156,"answer_zh":157,"source_url":158},2989,"如何安全地管理存储内容，防止意外覆盖？","系统采用不可变策略：一旦键（key）被添加，其值即视为不可修改。用户可以添加或删除键，但不能直接更新已有键的值。若需删除数据，请使用 remove 方法。责任委托给用户，操作需显式确认。","https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fissues\u002F14",{"id":160,"question_zh":161,"answer_zh":162,"source_url":163},2990,"是否支持从仪表板导出图表或数据？","支持多种导出格式。媒体文件支持 jpg, png, svg，导出格式包括 png, pdf, html。此外还支持导出 pandas dataframe、scikit-learn estimator、pandas series、numpy arrays 及原始数据类型。","https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fissues\u002F339",[165,170,175,180,185,190,195,200,205,210,215,220,225,230,235,240,245,250,255,260],{"id":166,"version":167,"summary_zh":168,"released_at":169},102494,"skore\u002F0.15.0","**Full Changelog**: https:\u002F\u002Fdocs.skore.probabl.ai\u002F0.15\u002Fchangelog.html","2026-04-02T09:34:45",{"id":171,"version":172,"summary_zh":173,"released_at":174},102495,"skore-mlflow-project\u002F0.0.4","**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fcompare\u002Fskore-mlflow-project\u002F0.0.3...skore-mlflow-project\u002F0.0.4","2026-04-02T10:03:12",{"id":176,"version":177,"summary_zh":178,"released_at":179},102496,"skore-local-project\u002F0.0.6","**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fcompare\u002Fskore-local-project\u002F0.0.5...skore-local-project\u002F0.0.6","2026-04-02T10:01:47",{"id":181,"version":182,"summary_zh":183,"released_at":184},102497,"skore-hub-project\u002F0.0.21","## What's Changed\r\n* feat(skore-hub-project): Dynamically load `SKORE_HUB_URI` by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2698\r\n* feat(skore-hub-project): Change the way the report cache is inspected for permutation importances by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2664\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fcompare\u002Fskore-hub-project\u002F0.0.20...skore-hub-project\u002F0.0.21","2026-04-02T08:43:58",{"id":186,"version":187,"summary_zh":188,"released_at":189},102498,"skore-hub-project\u002F0.0.20","## What's Changed\r\n* chore(skore-hub-project): Log in debug mode the response on error by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2645\r\n* feat(skore-hub-project)!: Revisit the cross-validation splitting strategy by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2609\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fcompare\u002Fskore-hub-project\u002F0.0.19...skore-hub-project\u002F0.0.20","2026-03-25T15:26:14",{"id":191,"version":192,"summary_zh":193,"released_at":194},102499,"skore-mlflow-project\u002F0.0.3","## What's Changed\r\n* fix(skore-mlflow-project): Handle heterogeneous datasets  by @cakedev0 in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2644\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fcompare\u002Fskore-mlflow-project\u002F0.0.2...skore-mlflow-project\u002F0.0.3","2026-03-24T10:45:06",{"id":196,"version":197,"summary_zh":198,"released_at":199},102500,"skore\u002F0.14.0","**Full Changelog**: https:\u002F\u002Fdocs.skore.probabl.ai\u002F0.14\u002Fchangelog.html#id2","2026-03-19T11:19:44",{"id":201,"version":202,"summary_zh":203,"released_at":204},102501,"skore-hub-project\u002F0.0.19","## What's Changed\r\n* fix(skore-hub-project): Compute confusion matrix with all thresholds by @rouk1 in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2590\r\n* fix(skore-hub-project): Add average precision and roc auc to binary classification performance by @rouk1 in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2605\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fcompare\u002Fskore-hub-project\u002F0.0.18...skore-hub-project\u002F0.0.19","2026-03-19T13:32:03",{"id":206,"version":207,"summary_zh":208,"released_at":209},102502,"skore-mlflow-project\u002F0.0.2","## What's Changed\r\n* chore(skore-mlflow-project): Ensure compatibility with https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2545 by @cakedev0 in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2580\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fcompare\u002Fskore-mlflow-project\u002F0.0.1...skore-mlflow-project\u002F0.0.2","2026-03-11T08:41:38",{"id":211,"version":212,"summary_zh":213,"released_at":214},102503,"skore-mlflow-project\u002F0.0.1","## What's Changed\r\n* feat(skore-mlflow-project): Add integration with MLflow by @cakedev0 in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2527\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fcommit\u002F015a0017a20fe0d1528093992e582b55ca4c785c","2026-03-06T11:57:26",{"id":216,"version":217,"summary_zh":218,"released_at":219},102504,"skore\u002F0.13.1","## What's Changed\r\n* fix(skore): Ensure figure-level legend is included when saving PredictionErrorDisplay plot by @MuditAtrey in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2530\r\n* feat(skore): Add aggregate parameter to ImpurityDecreaseDisplay by @MuditAtrey in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2539\r\n* refactor(skore): Calling underlying estimator method instead of factory by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2516\r\n* feat(skore\u002FComparisonReport): Add permutation importance by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2511\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fcompare\u002Fskore\u002F0.13.0...skore\u002F0.13.1","2026-03-05T10:21:19",{"id":221,"version":222,"summary_zh":223,"released_at":224},102505,"skore-hub-project\u002F0.0.18","## What's Changed\r\n* fix(skore-hub-project\u002Fsplitting_strategy): Use `_safe_indexing` with `report.y` by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2550\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fcompare\u002Fskore-hub-project\u002F0.0.17...skore-hub-project\u002F0.0.18","2026-03-03T13:53:39",{"id":226,"version":227,"summary_zh":228,"released_at":229},102512,"skore-hub-project\u002F0.0.15","## What's Changed\r\n* feat: Add target range by @rouk1 in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2271\r\n* feat: Send target distribution for regression tasks by @rouk1 in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2248\r\n* feat: Thread the compute of report metrics by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2178\r\n* fix(hubclient): Send `X-Skore-Client` with semver instead of PyPI version by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2243\r\n* fix(client): Fix bad import of `importlib.metadata` in conda by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2192\r\n* fix(login): Ignore token URI when using API key by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2191\r\n* feat!: Improve cross-validation splits property by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2214\r\n* feat: Add performance media to cross-validation report by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2215\r\n* feat!: Rename `splits` to `splitting_strategy` by @rouk1 in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2234\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fcompare\u002Fskore-hub-project\u002F0.0.13...skore-hub-project\u002F0.0.15","2026-01-12T14:01:23",{"id":231,"version":232,"summary_zh":233,"released_at":234},102513,"skore-hub-project\u002F0.0.15-rc.4","## What's Changed\r\n* feat: Add target range by @rouk1 in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2271\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fcompare\u002Fskore-hub-project\u002F0.0.15-rc.3...skore-hub-project\u002F0.0.15-rc.4","2026-01-09T14:02:16",{"id":236,"version":237,"summary_zh":238,"released_at":239},102506,"skore\u002F0.13.0","## What's Changed\r\n* fix(skore): Standardize dataframe column names to singular form by @Sharkyii in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2392\r\n* feat(skore): Add an HTML representation for interactive environment by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2316\r\n* fix(skore): Different y axis for comparisons in CoefficientsDisplay by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2403\r\n* chore(skore): Suppress seaborn deprecation warnings in box\u002Fstrip plots by @Sharkyii in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2405\r\n* fix(skore): Prevent memory overflow from dangerous arange by @cakedev0 in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2404\r\n* fix(skore\u002Fprogress-bar): Disable progress-bar on demand by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2425\r\n* chore(skore\u002Finspection): Increase alpha in shaded background by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2428\r\n* fix(skore\u002Freports): Fix auto-completion in `IPython` by @jeromedockes in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2427\r\n* chore(skore\u002Fcross-validation-report): Refine progress-bar message on split predictions by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2400\r\n* feat(skore): Allow to set pos_label after creating report by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2438\r\n* fix(skore): Don't crash when y is a list\u002Ftuple by @cakedev0 in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2433\r\n* fix(skore): Link to right repo in source url by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2448\r\n* feat(skore): Enable interactive `matplotlib` mode by default by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2446\r\n* docs(skore): Mention `split` in class docstring by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2447\r\n* fix(skore): Fix the link to source code in docs by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2451\r\n* feat(skore\u002FProject): Clarify single ML task constraint by @auguste-probabl in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2452\r\n* fix(skore): Don't create a tuple from a string for the cache key by @cakedev0 in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2456\r\n* fix(skore): Robustify normalization of X and y to a frame (`_retrieve_data_as_frame`) by @cakedev0 in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2440\r\n* fix(skore): Add support for multioutput regression in prediction error display by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2434\r\n* chore(skore): Make ipython and ipywidget optional dependencies by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2482\r\n* refactor(skore): Remove _verbose_name in favor of using registered named from accessor by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2475\r\n* fix(skore): Do not catch exception in cross-validation fitting by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2462\r\n* feat(skore\u002Fproject)!: Enforce `mode` in a dedicated parameter by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2401\r\n* docs(skore): Add small example to illustrate the usage of the skore.Project in local by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2481\r\n* fix(skore): Improve error message if missing skore-hub-project by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2458\r\n* refactor(skore): Introduce _report_type to avoid hard coding when needed by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2476\r\n* fix(skore): Fix wrong cache read \u002F missing check in metrics accessor (CV report) by @cakedev0 in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2492\r\n* feat(skore\u002FConfusionMatrixDislay): Introduce `threshold_value=\"all\"` option by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2463\r\n* chore(skore): Disallow clustering models in reports by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2489\r\n* feat(skore\u002FComparisonReport): Add `impurity_decrease` method by @auguste-probabl in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2387\r\n* refactor(skore): Uniform cache key sanitizing by @cakedev0 in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2483\r\n* fix(skore\u002Fconfiguration)!: Make the configuration inheritable between threads by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2431\r\n* fix(skore): Disable unecessary plotting made by `skrub` by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2494\r\n* fix(skore\u002Ftests): Make tests pass with `pandas>=3` by @auguste-probabl in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2504\r\n* feat(skore): Add cross-validation support for permutation importance by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2370\r\n* fix(skore): Fix the index in permutation importance and docs by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2506\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fcompare\u002Fskore\u002F0.12.0...skore\u002F0.13.0","2026-02-26T15:39:00",{"id":241,"version":242,"summary_zh":243,"released_at":244},102507,"skore-hub-project\u002F0.0.17","## What's Changed\r\n* fix(skore-hub-project): Replace `switch_mpl_backend` by `switch_mpl_ioff` by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2439\r\n* fix(skore-hub-project): Raise when project name is empty by @auguste-probabl in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2453\r\n* fix(skore-hub-project): Raise earlier an error when pos_label not provided before serialization by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2436\r\n* fix(skore-hub-project): Limit project name length to 64 by @auguste-probabl in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2454\r\n* fix(skore-hub-project): Serialize `pandas.Timestamp` by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2484\r\n* fix(skore-hub-project): Opt into pandas 3 behaviour by @auguste-probabl in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2485\r\n* feat(skore-hub-project\u002Fproject): Refactor initialization to raise custom exceptions by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2487\r\n* feat(skore-hub-project): Print report URL after put by @rouk1 in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2488\r\n* fix(skore-hub-project): Do not aggregate importance when sending to hub by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2445\r\n* fix(skore-hub-project): Restore `switch_mpl_backend` for use in conjunction with `plt.ioff()` by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2498\r\n* feat(skore-hub-project\u002Fput): Improve progress tracking by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2424\r\n* feat(skore-hub-project): Send inspection media for cross-validation reports by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2510\r\n* feat(skore-hub-project\u002Fput): Send also performance media as dataframe by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2514\r\n* feat(skore-hub-project\u002Fput): Send confusion matrix by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2525\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fcompare\u002Fskore-hub-project\u002F0.0.16...skore-hub-project\u002F0.0.17","2026-02-26T15:49:35",{"id":246,"version":247,"summary_zh":248,"released_at":249},102508,"skore\u002F0.12.0","## What's Changed\r\n* feat(skore): Add support for ComparisonReport in ConfusionMatrixDisplay by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2236\r\n* fix(skore): Use display API for coefficients display by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2152\r\n* chore(skore): Moving confusion matrix display kwargs into set_style by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2293\r\n* chore(skore): Clean up docs references to seaborn by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2294\r\n* fix: Add accuracy as default metric for classifiers by @auguste-probabl in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2295\r\n* chore(skore): Moving precision recall kwargs into set_style by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2298\r\n* chore(skore): Moving coefficients kwargs into set_style by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2300\r\n* chore(skore): Moving table report kwargs into set_style by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2299\r\n* chore(skore): Improve documentation of set_style() by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2306\r\n* feat(skore): Refactor PredictionErrorDisplay to use seaborn and add subplot_by option by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2292\r\n* fix(skore): Actually suppress warning in prc display by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2312\r\n* feat!: Rename `indicator_favorability` to `favorability` by @auguste-probabl in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2314\r\n* chore(skore): Add the feature importance accessor to the API reference by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2315\r\n* feat(skore): Refactor RocCurveDisplay to use seaborn and add subplot_by by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2313\r\n* feat(skore): Rename `tenant` to `workspace` by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2325\r\n* chore(skore): Make summary and widget modules private by @ShantKhatri in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2332\r\n* feat(skore): Create `PermutationImportanceDisplay` for `permutation()` method by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2304\r\n* feat(skore): Create an EstimatorReport from a Comparison\u002FCrossValidationReport by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2326\r\n* feat(skore): Add data_source=\"both\" to ComparisonReport metrics frames by @Adrien-1997 in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2201\r\n* fix(skore): Cast report to `EstimatorReport` for mypy  by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2345\r\n* feat(skore\u002FCoefficientsDisplay): Add `select_k` and `sorting_order` parameters to `frame` and `plot` by @adwait-godbole in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2305\r\n* chore(skore): Homogenize display factories & follow up datasource='both' by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2343\r\n* chore(skore)!: Rename `feature_importance` to `inspection` and the underlying functions by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2346\r\n* feat(skore)!: Refactor hub authentication by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2340\r\n* fix(skore): Error with set_style in PermutationImportanceDisplay by @Premkumar-2004 in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2364\r\n* chore(skore)!: Homogenize the attributes in Displays by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2367\r\n* feat(skore): Create ImpurityDecreaseDisplay for the `impurity_decrease` method by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2341\r\n* fix(skore\u002FCoefficientsDisplay): Change default ordering to `None` by @Premkumar-2004 in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2375\r\n* fix(skore): Avoid accessing frame[output] when the column does not exist by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2382\r\n* fix(skore): Make the progress thread-safe by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2357\r\n* chore(skore\u002Festimator-report): Clean some parts of the metrics accessor by @auguste-probabl in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2331\r\n* feat(skore\u002FCrossValidationReport): Add `impurity_decrease` method by @auguste-probabl in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2385\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fcompare\u002Fskore\u002F0.11.7...skore\u002F0.12.0","2026-02-09T13:18:45",{"id":251,"version":252,"summary_zh":253,"released_at":254},102509,"skore-hub-project\u002F0.0.16","## What's Changed\r\n* ci(skore-hub-project): Filter deprecation warnings caused by `seaborn` when using `skore>=0.11.6` by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2282\r\n* fix(skore-hub-project\u002Fmetrics): Raise exception from thread by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2296\r\n* fix(skore-hub-project\u002Ftests): Ignore `Precision is ill-defined` warning by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2307\r\n* chore(skore-hub-project\u002Ftests): Improve `test_metrics_raises_exception` by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2311\r\n* fix(skore-hub-project\u002Fproject): Slugify tenant and project name by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2321 and https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2339\r\n* feat(skore-hub-project): Rename `tenant` to `workspace` by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2324\r\n* feat(skore-hub-project)!: Refactor authentication by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2333\r\n* feat(skore-hub-project)!: Ensure compatibility with `skore>=0.12` by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2378\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fcompare\u002Fskore-hub-project\u002F0.0.15...skore-hub-project\u002F0.0.16","2026-02-09T13:37:53",{"id":256,"version":257,"summary_zh":258,"released_at":259},102510,"skore\u002F0.11.7","## What's Changed\r\n* fix: Allow subplot_by=None for PRC in multiclass by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2273\r\n* docs: Rework landing page to show takeaway messages by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2281\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fcompare\u002Fskore\u002F0.11.6...skore\u002F0.11.7","2026-01-14T09:27:39",{"id":261,"version":262,"summary_zh":263,"released_at":264},102511,"skore\u002F0.11.6","## What's Changed\r\n\r\n### Chore\r\n\r\n* docs: Use `skrub` instead of `scikit-learn` to fetch california housing dataset by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2244\r\n* ci: Drop support of `scikit-learn==1.4` in favor of `scikit-learn==1.8` by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2209\r\n* docs: Use `reo.dev` tracker on all pages by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2246\r\n* chore: Refresh dates in LICENSE by @auguste-probabl in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2263\r\n* docs: Specify explicitly the subplot_by available string by @glemaitre in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2272\r\n* docs: Remove large dependencies by @auguste-probabl in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2270\r\n* docs: Refresh \"getting started\" example by @auguste-probabl in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2269\r\n* docs: Replace `plausible` by `posthog` by @thomass-dev in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2279\r\n\r\n### Improvements\r\n\r\n* feat: Refactor PR curve display and add `subplot_by` option by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2193\r\n* feat: Add support for CrossValidationReport in ConfusionMatrixDisplay by @GaetandeCast in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2221\r\n* fix: Add `response_method` argument to every metrics accessor by @auguste-probabl in https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fpull\u002F2276\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fprobabl-ai\u002Fskore\u002Fcompare\u002Fskore\u002F0.11.5...skore\u002F0.11.6","2026-01-13T09:13:26"]