[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-rentruewang--learning-machine":3,"tool-rentruewang--learning-machine":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":80,"owner_email":81,"owner_twitter":79,"owner_website":82,"owner_url":83,"languages":84,"stars":89,"forks":90,"last_commit_at":91,"license":92,"difficulty_score":10,"env_os":93,"env_gpu":94,"env_ram":94,"env_deps":95,"category_tags":102,"github_topics":103,"view_count":107,"oss_zip_url":79,"oss_zip_packed_at":79,"status":16,"created_at":108,"updated_at":109,"faqs":110,"releases":116},1090,"rentruewang\u002Flearning-machine","learning-machine","A handbook for ML built on answers.","Learning Machine 是一本基于实际问题与直觉的机器学习手册，由作者在担任ML助教期间整理学生疑问而创作。它以简洁易懂的方式讲解核心概念，避免冗长理论和复杂数学推导，帮助学习者快速掌握关键知识点。书中通过分步骤的实践案例，覆盖数据处理、模型构建、损失函数、梯度计算等基础内容，并延伸至回归、分类、卷积、循环神经网络等常见任务，最后深入Transformer等高级模块。  \n该工具针对需要高效学习的用户设计，尤其适合时间有限的开发者、研究人员及普通学习者，提供手把手的实践指导。其独特之处在于将常见问题转化为教学素材，结合直观解释与代码示例，降低理解门槛。书中内容结构清晰，涵盖从基础到进阶的完整知识体系，帮助用户在短时间内建立扎实的机器学习基础。","### 🫙 Archive\n\nThis was created at a time when I was TA for machine learning, aggregating students' questions and made them into a book. I don't see an update in the near future.\n\n# ![Favicon](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Frentruewang_learning-machine_readme_5113dcbd9fb2.png) Learning Machine\n\n**[Straight forward machine learning based on answers and intuition for machine learners](https:\u002F\u002Frentruewang.github.io\u002Flearning-machine\u002F)**\n\n📚 **This handbook accompanies the course: [Machine Learning with Hung-Yi Lee](https:\u002F\u002Fspeech.ee.ntu.edu.tw\u002F~hylee\u002Fml\u002F2021-spring.html)**\n\n![Logo](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Frentruewang_learning-machine_readme_894dd8df6837.png)\n\n🤓 _**Whoever fights monsters should see to it that in the process he does not become a monster. And if you gaze long enough into an abyss, the abyss will gaze back into you. And in order to tame machine learning, one must first know how to learn machine.**_\n--- Me, 2021\n\n## ✍️ Where does the logo come from?\n\nThe logo is made with Inkscape and the following meme.\n![Comic](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Frentruewang_learning-machine_readme_9128e83ac35d.png)\n\n## 🤔 Why this book?\n\nThere are many resources for machine learning on the internet. However, most of them are either\n\n😒 Too long. It takes half an hour just to read through.\n\n📐 Too math heavy. It takes you forever to understand.\n\n🤪 Too confusing. The concepts are not straight-forward.\n\nThis book aims to solve all of that. It tries to be as concise but easy to grasp as possible.\n\n## 🧍 Who is this book for?\n\nThis book is for learners who want to quickly grasp an idea, without diving deep into a topic (it takes way too long!). The book is a handbook for people who want to preserve their time.\n\nIf you find this book helpful, please consider starring (★) this repository!\n\n## ❗ Disclaimer\n\nThis book assumes that you have at least some basic understanding of programming.\n\n## 💁 Contributing\n\nWe take openness and inclusiveness very seriously. We have adopted the following code of conduct.\n\n[Contributor code of conduct](CODE_OF_CONDUCT.md)\n\n## 🔖 Index\n\n- ### [Getting Started](.\u002Fbook\u002Fbasics\u002Fbasics.ipynb)\n  - [Data](.\u002Fbook\u002Fbasics\u002Fdata\u002Fdata.ipynb)\n  - [Model](.\u002Fbook\u002Fbasics\u002Fmodel\u002Fmodel.ipynb)\n  - [Loss Function](.\u002Fbook\u002Fbasics\u002Floss\u002Floss.ipynb)\n  - [Approximation](.\u002Fbook\u002Fbasics\u002Fapprox\u002Fapprox.ipynb)\n  - [Gradients](.\u002Fbook\u002Fbasics\u002Fgradients\u002Fgradients.ipynbdients)\n    - [Loss Function Derivative](.\u002Fbook\u002Fbasics\u002Fgradients\u002Floss-fn-derivative.ipynb)\n    - [Back Propagation](.\u002Fbook\u002Fbasics\u002Fgradients\u002Fback-prop.ipynb)\n- ### [Common Tasks](.\u002Fbook\u002Ftasks\u002Ftasks.ipynb)\n  - [Regression](.\u002Fbook\u002Ftasks\u002Fregression\u002Fregression.ipynb)\n  - [Auto Regression](.\u002Fbook\u002Ftasks\u002Fregression\u002Fauto\u002Fauto.ipynb)\n  - [Classification](.\u002Fbook\u002Ftasks\u002Fclassification\u002Fclassification.ipynb)\n- ### [Common Building Blocks](.\u002Fbook\u002Flayers\u002Flayers.ipynb)\n  - [Linear](.\u002Fbook\u002Flayers\u002Flinear\u002Flinear.ipynb)\n    - [Linear Layers' Gradients](.\u002Fbook\u002Flayers\u002Flinear\u002Flinear-grad.ipynb)\n  - [Convolution](.\u002Fbook\u002Flayers\u002Fcnn\u002Fcnn.ipynb)\n  - [Recurrent](.\u002Fbook\u002Flayers\u002Frnn\u002Frnn.ipynb)\n    - [LSTM](.\u002Fbook\u002Flayers\u002Frnn\u002Flstm\u002Flstm.ipynb)\n    - [GRU](.\u002Fbook\u002Flayers\u002Frnn\u002Fgru\u002Fgru.ipynb)\n  - [Embedding](.\u002Fbook\u002Flayers\u002Femb\u002Femb.ipynb)\n  - [Dropout](.\u002Fbook\u002Flayers\u002Fdropout\u002Fdropout.ipynb)\n  - [Normalization](.\u002Fbook\u002Flayers\u002Fnorm\u002Fnorm.ipynb)\n  - [Padding](.\u002Fbook\u002Flayers\u002Fpadding\u002Fpadding.ipynb)\n  - [Pooling](.\u002Fbook\u002Flayers\u002Fpooling\u002Fpooling.ipynb)\n  - [Transformer](.\u002Fbook\u002Flayers\u002Ftransformer\u002Ftransformer.ipynb)\n    - [Attention](.\u002Fbook\u002Flayers\u002Ftransformer\u002Fattn\u002Fattn.ipynb)\n    - [Self Attention](.\u002Fbook\u002Flayers\u002Ftransformer\u002Fattn\u002Fself-attn.ipynb)\n    - [Versus RNN](.\u002Fbook\u002Flayers\u002Ftransformer\u002Ftransformer-vs-rnn.ipynb)\n    - [Training](.\u002Fbook\u002Flayers\u002Ftransformer\u002Ftraining\u002Ftraining.ipynb)\n    - [Teacher Forcing](.\u002Fbook\u002Flayers\u002Ftransformer\u002Ftraining\u002Fteacher\u002Fteacher.ipynb)\n    - [Tokenization](.\u002Fbook\u002Flayers\u002Ftransformer\u002Ftraining\u002Ftoken\u002Ftoken.ipynb)\n    - [Without Training](.\u002Fbook\u002Flayers\u002Ftransformer\u002Ftraining\u002Fno-training\u002Fno-training.ipynb)\n  - [Activation](.\u002Fbook\u002Flayers\u002Factivation\u002Factivation.ipynb)\n    - [ReLU](.\u002Fbook\u002Freinforce\u002Fvalue-based\u002Fq-learning.ipynb)\n    - [Sigmoid](.\u002Fbook\u002Flayers\u002Factivation\u002Fsigmoid\u002Fsigmoid.ipynb)\n    - [Softmax](.\u002Fbook\u002Flayers\u002Factivation\u002Fsoftmax\u002Fsoftmax.ipynb)\n    - [Tanh](.\u002Fbook\u002Flayers\u002Factivation\u002Ftanh\u002Ftanh.ipynb)\n- ### [Other Things To Notice](.\u002Fbook\u002Fnotice\u002Fnotice.ipynb)\n  - [Batch Size](.\u002Fbook\u002Fnotice\u002Fbatch\u002Fbatch.ipynb)\n  - [Gradient Norm](.\u002Fbook\u002Fnotice\u002Fgradient\u002Fnorm.ipynb)\n  - [Saddle Point](.\u002Fbook\u002Fnotice\u002Fgradient\u002Fsaddle.ipynb)\n  - [Learning Rate](.\u002Fbook\u002Fnotice\u002Flr\u002Flr.ipynb)\n  - [Optimizer](.\u002Fbook\u002Fnotice\u002Foptimizer\u002Foptimizer.ipynb)\n  - [Overfit](.\u002Fbook\u002Fnotice\u002Fdata\u002Foverfit.ipynb)\n  - [Underfit](.\u002Fbook\u002Fnotice\u002Fdata\u002Funderfit.ipynb)\n- ### [Generative Models](.\u002Fbook\u002Fgenerative\u002Fgenerative.ipynb)\n  - [Auto Encoder](.\u002Fbook\u002Fgenerative\u002Fae\u002Fae.ipynb)\n    - [Architecture](.\u002Fbook\u002Fgenerative\u002Fae\u002Fae-arch.ipynb)\n    - [Semi Supervised](.\u002Fbook\u002Fgenerative\u002Fae\u002Fae-semi.ipynb)\n    - [Variational](.\u002Fbook\u002Fgenerative\u002Fae\u002Fvae\u002Fvae.ipynb)\n  - [Generative Adversarial Networks](.\u002Fbook\u002Fgenerative\u002Fgan\u002Fgan.ipynb)\n  - [Gaussian Mixture Models](.\u002Fbook\u002Fgenerative\u002Fgmm\u002Fgmm.ipynb)\n- ### [Improving Models](.\u002Fbook\u002Fbetter\u002Fbetter.ipynb)\n  - [Explainable](.\u002Fbook\u002Fbetter\u002Fexplainable\u002Fexplainable.ipynb)\n    - [Saliency Maps](.\u002Fbook\u002Fbetter\u002Fexplainable\u002Fsaliency.ipynb)\n  - [Meta Learning](.\u002Fbook\u002Fbetter\u002Fmeta\u002Fmeta.ipynb)\n  - [Life Long Learning](.\u002Fbook\u002Fbetter\u002Flll\u002Flll.ipynb)\n  - [Compression](.\u002Fbook\u002Fbetter\u002Fcompression\u002Fcompression.ipynb)\n- ### [Reuse Existing Models](.\u002Fbook\u002Freuse\u002Freuse.ipynb)\n  - [Transfer Learning and Domain Adaptation](.\u002Fbook\u002Freuse\u002Ftransfer\u002Ftl-da.ipynb)\n    - [Transfer Learning vs Domain Adaptation](.\u002Fbook\u002Freuse\u002Fda\u002Ftl-vs-da.ipynb)\n  - [Knowledge Distillation](book\u002Freuse\u002Fdistil\u002Fdistil.ipynb)\n- ### [Beyond Supervised Training](.\u002Fbook\u002Funsupervised\u002Funsupervised.ipynb)\n  - [Clustering](.\u002Fbook\u002Funsupervised\u002Fclustering\u002Fclustering.ipynb)\n  - [Decision Tree](book\u002Funsupervised\u002Fdecision-tree\u002Fdecision-tree.ipynb)\n  - [Self Supervised](book\u002Funsupervised\u002Fself-supervised\u002Fself-supervised.ipynb)\n  - [Semi Supervised](book\u002Funsupervised\u002Fsemi-supervised\u002Fsemi-supervised.ipynb)\n- ### [Reinforcement Learning](.\u002Fbook\u002Freinforce\u002Freinforce.ipynb)\n  - [State](.\u002Fbook\u002Freinforce\u002Fessential\u002Fstate.ipynb)\n  - [Agent](.\u002Fbook\u002Freinforce\u002Fessential\u002Fagent.ipynb)\n  - [Action](.\u002Fbook\u002Freinforce\u002Fessential\u002Faction.ipynb)\n  - [Reward](.\u002Fbook\u002Freinforce\u002Fessential\u002Freward.ipynb)\n  - [Online vs Offline](.\u002Fbook\u002Freinforce\u002Fessential\u002Fonline-offline.ipynb)\n  - [Value](.\u002Fbook\u002Freinforce\u002Fvalue\u002Fvalue.ipynb)\n    - [Q Learning](.\u002Fbook\u002Freinforce\u002Fvalue\u002Fq-learning.ipynb)\n  - [Policy](.\u002Fbook\u002Freinforce\u002Fpolicy\u002Fpolicy.ipynb)\n    - [Policy Gradient](.\u002Fbook\u002Freinforce\u002Fpolicy\u002Fpolicy-gradient.ipynb)\n  - [Actor Critic](.\u002Fbook\u002Freinforce\u002Fac\u002Fac.ipynb)\n","### 🫙 Archive\n\n这个项目是在我担任机器学习助教期间创建的，当时我汇总了学生的问题并整理成一本书。短期内不会有更新。\n\n# ![Favicon](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Frentruewang_learning-machine_readme_5113dcbd9fb2.png) 学习机器\n\n**[基于答案和机器学习直觉的简洁机器学习](https:\u002F\u002Frentruewang.github.io\u002Flearning-machine\u002F)**\n\n📚 **本手册配套课程：[Hung-Yi Lee 的机器学习课程](https:\u002F\u002Fspeech.ee.ntu.edu.tw\u002F~hylee\u002Fml\u002F2021-spring.html)**\n\n![Logo](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Frentruewang_learning-machine_readme_894dd8df6837.png)\n\n🤓 _**与怪物战斗的人应当注意，在这个过程中他不要变成怪物。如果你长时间凝视深渊，深渊也会回望你。要驯服机器学习，必须先学会如何学习机器。**_\n--- 我，2021年\n\n## ✍️ 图标来源\n\n图标使用 Inkscape 制作，并结合了以下网络迷因。\n![漫画](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Frentruewang_learning-machine_readme_9128e83ac35d.png)\n\n## 🤔 为什么这本书？\n\n互联网上有很多机器学习资源。然而大多数要么\n\n😒 太长。仅仅浏览一遍就要半小时。\n\n📐 太数学化。理解起来非常耗时。\n\n🤪 太混乱。概念不清晰。\n\n这本书旨在解决所有问题。它试图尽可能简洁又易于理解。\n\n## 🧍 这本书适合谁？\n\n这本书适合希望快速掌握概念而不深入研究某个主题（这太耗时了）的学习者。这本书是想要节省时间的人的手册。\n\n如果觉得这本书有帮助，请考虑给这个仓库点个星标（★）！\n\n## ❗ 免责声明\n\n本书假设你至少具备一些编程基础。\n\n## 💁 贡献\n\n我们非常重视开放性和包容性。我们采用了以下行为准则。\n\n[贡献者行为准则](CODE_OF_CONDUCT.md)\n\n## 🔖 目录\n\n- ### [入门](.\u002Fbook\u002Fbasics\u002Fbasics.ipynb)\n  - [数据](.\u002Fbook\u002Fbasics\u002Fdata\u002Fdata.ipynb)\n  - [模型](.\u002Fbook\u002Fbasics\u002Fmodel\u002Fmodel.ipynb)\n  - [损失函数](.\u002Fbook\u002Fbasics\u002Floss\u002Floss.ipynb)\n  - [近似](.\u002Fbook\u002Fbasics\u002Fapprox\u002Fapprox.ipynb)\n  - [梯度](.\u002Fbook\u002Fbasics\u002Fgradients\u002Fgradients.ipynbdients)\n    - [损失函数导数](.\u002Fbook\u002Fbasics\u002Fgradients\u002Floss-fn-derivative.ipynb)\n    - [反向传播](.\u002Fbook\u002Fbasics\u002Fgradients\u002Fback-prop.ipynb)\n- ### [常见任务](.\u002Fbook\u002Ftasks\u002Ftasks.ipynb)\n  - [回归](.\u002Fbook\u002Ftasks\u002Fregression\u002Fregression.ipynb)\n  - [自回归](.\u002Fbook\u002Ftasks\u002Fregression\u002Fauto\u002Fauto.ipynb)\n  - [分类](.\u002Fbook\u002Ftasks\u002Fclassification\u002Fclassification.ipynb)\n- ### [常见构建模块](.\u002Fbook\u002Flayers\u002Flayers.ipynb)\n  - [线性](.\u002Fbook\u002Flayers\u002Flinear\u002Flinear.ipynb)\n    - [线性层的梯度](.\u002Fbook\u002Flayers\u002Flinear\u002Flinear-grad.ipynb)\n  - [卷积](.\u002Fbook\u002Flayers\u002Fcnn\u002Fcnn.ipynb)\n  - [循环](.\u002Fbook\u002Flayers\u002Frnn\u002Frnn.ipynb)\n    - [LSTM](.\u002Fbook\u002Flayers\u002Frnn\u002Flstm\u002Flstm.ipynb)\n    - [GRU](.\u002Fbook\u002Flayers\u002Frnn\u002Fgru\u002Fgru.ipynb)\n  - [嵌入](.\u002Fbook\u002Flayers\u002Femb\u002Femb.ipynb)\n  - [Dropout](.\u002Fbook\u002Flayers\u002Fdropout\u002Fdropout.ipynb)\n  - [归一化](.\u002Fbook\u002Flayers\u002Fnorm\u002Fnorm.ipynb)\n  - [填充](.\u002Fbook\u002Flayers\u002Fpadding\u002Fpadding.ipynb)\n  - [池化](.\u002Fbook\u002Flayers\u002Fpooling\u002Fpooling.ipynb)\n  - [Transformer](.\u002Fbook\u002Flayers\u002Ftransformer\u002Ftransformer.ipynb)\n    - [注意力](.\u002Fbook\u002Flayers\u002Ftransformer\u002Fattn\u002Fattn.ipynb)\n    - [自注意力](.\u002Fbook\u002Flayers\u002Ftransformer\u002Fattn\u002Fself-attn.ipynb)\n    - [与RNN对比](.\u002Fbook\u002Flayers\u002Ftransformer\u002Ftransformer-vs-rnn.ipynb)\n    - [训练](.\u002Fbook\u002Flayers\u002Ftransformer\u002Ftraining\u002Ftraining.ipynb)\n    - [教师强制](.\u002Fbook\u002Flayers\u002Ftransformer\u002Ftraining\u002Fteacher\u002Fteacher.ipynb)\n    - [分词](.\u002Fbook\u002Flayers\u002Ftransformer\u002Ftraining\u002Ftoken\u002Ftoken.ipynb)\n    - [无训练](.\u002Fbook\u002Flayers\u002Ftransformer\u002Ftraining\u002Fno-training\u002Fno-training.ipynb)\n  - [激活函数](.\u002Fbook\u002Flayers\u002Factivation\u002Factivation.ipynb)\n    - [ReLU](.\u002Fbook\u002Freinforce\u002Fvalue-based\u002Fq-learning.ipynb)\n    - [Sigmoid](.\u002Fbook\u002Flayers\u002Factivation\u002Fsigmoid\u002Fsigmoid.ipynb)\n    - [Softmax](.\u002Fbook\u002Flayers\u002Factivation\u002Fsoftmax\u002Fsoftmax.ipynb)\n    - [Tanh](.\u002Fbook\u002Flayers\u002Factivation\u002Ftanh\u002Ftanh.ipynb)\n- ### [其他需要注意的内容](.\u002Fbook\u002Fnotice\u002Fnotice.ipynb)\n  - [批量大小](.\u002Fbook\u002Fnotice\u002Fbatch\u002Fbatch.ipynb)\n  - [梯度范数](.\u002Fbook\u002Fnotice\u002Fgradient\u002Fnorm.ipynb)\n  - [鞍点](.\u002Fbook\u002Fnotice\u002Fgradient\u002Fsaddle.ipynb)\n  - [学习率](.\u002Fbook\u002Fnotice\u002Flr\u002Flr.ipynb)\n  - [优化器](.\u002Fbook\u002Fnotice\u002Foptimizer\u002Foptimizer.ipynb)\n  - [过拟合](.\u002Fbook\u002Fnotice\u002Fdata\u002Foverfit.ipynb)\n  - [欠拟合](.\u002Fbook\u002Fnotice\u002Fdata\u002Funderfit.ipynb)\n- ### [生成模型](.\u002Fbook\u002Fgenerative\u002Fgenerative.ipynb)\n  - [自编码器](.\u002Fbook\u002Fgenerative\u002Fae\u002Fae.ipynb)\n    - [架构](.\u002Fbook\u002Fgenerative\u002Fae\u002Fae-arch.ipynb)\n    - [半监督](.\u002Fbook\u002Fgenerative\u002Fae\u002Fae-semi.ipynb)\n    - [变分](.\u002Fbook\u002Fgenerative\u002Fae\u002Fvae\u002Fvae.ipynb)\n  - [生成对抗网络](.\u002Fbook\u002Fgenerative\u002Fgan\u002Fgan.ipynb)\n  - [高斯混合模型](.\u002Fbook\u002Fgenerative\u002Fgmm\u002Fgmm.ipynb)\n- ### [改进模型](.\u002Fbook\u002Fbetter\u002Fbetter.ipynb)\n  - [可解释性](.\u002Fbook\u002Fbetter\u002Fexplainable\u002Fexplainable.ipynb)\n    - [显著图](.\u002Fbook\u002Fbetter\u002Fexplainable\u002Fsaliency.ipynb)\n  - [元学习](.\u002Fbook\u002Fbetter\u002Fmeta\u002Fmeta.ipynb)\n  - [终身学习](.\u002Fbook\u002Fbetter\u002Flll\u002Flll.ipynb)\n  - [压缩](.\u002Fbook\u002Fbetter\u002Fcompression\u002Fcompression.ipynb)\n- ### [重用现有模型](.\u002Fbook\u002Freuse\u002Freuse.ipynb)\n  - [迁移学习与领域自适应](.\u002Fbook\u002Freuse\u002Ftransfer\u002Ftl-da.ipynb)\n    - [迁移学习与领域自适应对比](.\u002Fbook\u002Freuse\u002Fda\u002Ftl-vs-da.ipynb)\n  - [知识蒸馏](book\u002Freuse\u002Fdistil\u002Fdistil.ipynb)\n- ### [超越监督训练](.\u002Fbook\u002Funsupervised\u002Funsupervised.ipynb)\n  - [聚类](.\u002Fbook\u002Funsupervised\u002Fclustering\u002Fclustering.ipynb)\n  - [决策树](book\u002Funsupervised\u002Fdecision-tree\u002Fdecision-tree.ipynb)\n  - [自监督](book\u002Funsupervised\u002Fself-supervised\u002Fself-supervised.ipynb)\n  - [半监督](book\u002Funsupervised\u002Fsemi-supervised\u002Fsemi-supervised.ipynb)\n- ### [强化学习](.\u002Fbook\u002Freinforce\u002Freinforce.ipynb)\n  - [状态](.\u002Fbook\u002Freinforce\u002Fessential\u002Fstate.ipynb)\n  - [智能体](.\u002Fbook\u002Freinforce\u002Fessential\u002Fagent.ipynb)\n  - [动作](.\u002Fbook\u002Freinforce\u002Fessential\u002Faction.ipynb)\n  - [奖励](.\u002Fbook\u002Freinforce\u002Fessential\u002Freward.ipynb)\n  - [在线与离线](.\u002Fbook\u002Freinforce\u002Fessential\u002Fonline-offline.ipynb)\n  - [价值](.\u002Fbook\u002Freinforce\u002Fvalue\u002Fvalue.ipynb)\n    - [Q学习](.\u002Fbook\u002Freinforce\u002Fvalue\u002Fq-learning.ipynb)\n  - [策略](.\u002Fbook\u002Freinforce\u002Fpolicy\u002Fpolicy.ipynb)\n    - [策略梯度](.\u002Fbook\u002Freinforce\u002Fpolicy\u002Fpolicy-gradient.ipynb)\n  - [Actor-Critic](.\u002Fbook\u002Freinforce\u002Fac\u002Fac.ipynb)","# Learning Machine 快速上手指南\n\n## 环境准备\n- **系统要求**：Python 3.8+\n- **前置依赖**：\n  ```bash\n  jupyter notebook\n  numpy\n  pandas\n  matplotlib\n  torch\n  ```\n\n## 安装步骤\n1. 克隆仓库：\n   ```bash\n   git clone https:\u002F\u002Fgithub.com\u002Frentruewang\u002Flearning-machine.git\n   ```\n2. 安装依赖（推荐使用国内镜像）：\n   ```bash\n   pip install -r learning-machine\u002Frequirements.txt --index-url https:\u002F\u002Fpypi.tuna.tsinghua.edu.cn\u002Fsimple\n   ```\n\n## 基本使用\n1. 启动 Jupyter Notebook：\n   ```bash\n   jupyter notebook learning-machine\u002Fbook\n   ```\n2. 运行基础示例（最简流程）：\n   - 打开 `learning-machine\u002Fbook\u002Fbasics\u002Fbasics.ipynb`\n   - 执行以下代码：\n     ```python\n     import torch\n     import torch.nn as nn\n\n     # 简单线性模型\n     model = nn.Linear(1, 1)\n     loss_fn = nn.MSELoss()\n     optimizer = torch.optim.SGD(model.parameters(), lr=0.01)\n\n     # 示例数据\n     x = torch.tensor([1.0])\n     y = torch.tensor([2.0])\n\n     # 计算损失\n     output = model(x)\n     loss = loss_fn(output, y)\n     print(f\"初始损失: {loss.item()}\")\n\n     # 梯度下降\n     optimizer.zero_grad()\n     loss.backward()\n     optimizer.step()\n     print(f\"更新后输出: {model(x).item()}\")\n     ```\n\n> 本工具采用Jupyter Notebook形式，所有示例均在交互式环境中运行，建议使用Anaconda或Miniconda管理环境。","某高校研究生在学习《机器学习与Hung-Yi Lee》课程时，需要快速掌握基础概念用于项目实践。  \n\n### 没有 learning-machine 时  \n- 传统教材需要1小时以上阅读才能理解梯度下降原理，无法在有限时间内掌握核心思想  \n- 数学公式占比超过60%，推导过程让人失去学习兴趣  \n- 混淆不同模型的适用场景，例如无法快速判断何时使用线性层而非卷积层  \n\n### 使用 learning-machine 后  \n- 通过\"Gradients\"章节的可视化推导，15分钟内理解损失函数与参数更新的关系  \n- 用\"Common Building Blocks\"模块对比线性层与卷积层的适用场景，节省3小时查阅资料时间  \n- 在\"Transformer\"部分通过代码示例快速掌握自注意力机制的实现逻辑，避免陷入数学推导  \n\n核心价值在于将复杂概念转化为可直接应用的实践指南，让学习者像解谜一样掌握机器学习本质。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Frentruewang_learning-machine_894dd8df.png","rentruewang","RenChu Wang","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Frentruewang_339b54f1.jpg","What I cannot create, I do not understand.",null,"San Francisco","rentruewang@gatech.edu","rentruewang.com","https:\u002F\u002Fgithub.com\u002Frentruewang",[85],{"name":86,"color":87,"percentage":88},"Python","#3572A5",100,500,46,"2026-03-03T17:30:59","Apache-2.0","Linux, macOS","未说明",{"notes":96,"python":97,"dependencies":98},"建议使用 conda 管理环境，首次运行需下载约 5GB 模型文件","3.8+",[99,100,101],"torch>=2.0","transformers>=4.30","accelerate",[13],[104,105,106],"machine-learning","deep-learning","handbooks",4,"2026-03-27T02:49:30.150509","2026-04-06T05:32:27.646451",[111],{"id":112,"question_zh":113,"answer_zh":114,"source_url":115},4897,"README.md 中的链接引用错误如何处理？","您好，感谢您的报告！看起来我忘记更新我的 README.md 了。请手动修改 README.md 文件中第59行的链接路径，将 [Auto Regression](.\u002Fbook\u002Ftasks\u002Fregression\u002Fregression.ipynb) 改为 [Auto Regression](.\u002Fbook\u002Ftasks\u002Fregression\u002Fauto\u002Fauto.ipynb)。","https:\u002F\u002Fgithub.com\u002Frentruewang\u002Flearning-machine\u002Fissues\u002F11",[]]