Tutorial - 标签 - mywebsitehttps://steven-yl.github.io/mywebsite/tags/tutorial/Tutorial - 标签 - mywebsiteHugo -- gohugo.iozh-CNsteven@gmail.com (Steven)steven@gmail.com (Steven)This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.Sun, 22 Mar 2026 12:00:00 +0800机械臂控制方法https://steven-yl.github.io/mywebsite/arm_control/Sun, 22 Mar 2026 12:00:00 +0800Stevenhttps://github.com/steven-ylhttps://steven-yl.github.io/mywebsite/arm_control/从位置/速度/力/导纳/阻抗总览出发，展开阻抗与导纳对比、期望轨迹与力控实现、末端与关节力检测、重力/摩擦/惯性估计、零空间融合及拖动示教，附公式与工程要点Flow Matching Guide and Code: Discrete Flow Matchinghttps://steven-yl.github.io/mywebsite/flow-matching-guide-and-code-discrete-flow-matching/Thu, 05 Mar 2026 00:00:00 +0000Stevenhttps://github.com/steven-ylhttps://steven-yl.github.io/mywebsite/flow-matching-guide-and-code-discrete-flow-matching/我们来详细整理并解释这段关于连续时间马尔可夫链（CTMC）的内容，使其更易于理解。

CTMC 是一种用于生成离散数据（比如文本、类别数据）的模型。你可以把它想象成一个在有限个离散状态之间随时间跳转的“粒子”，它按照一定的“速率”从一个状态跳到另一个状态。这与之前讨论的“流模型”（用于连续数据，如图像）形成对比，CTMC 是后续“离散流匹配”模型的基础。

]]>Flow Matching Guide and Code(项目解析)https://steven-yl.github.io/mywebsite/flow-matching-guide-and-code%E9%A1%B9%E7%9B%AE%E8%A7%A3%E6%9E%90/Sat, 28 Feb 2026 19:37:39 +0800Stevenhttps://github.com/steven-ylhttps://steven-yl.github.io/mywebsite/flow-matching-guide-and-code%E9%A1%B9%E7%9B%AE%E8%A7%A3%E6%9E%90/<div class="featured-image"> <img src="/mywebsite/posts/images/flow-matching-guide-and-code-%e9%a1%b9%e7%9b%ae%e8%a7%a3%e6%9e%90.webp" referrerpolicy="no-referrer"> </div>Meta flow_matching 库与论文《Flow Matching Guide and Code》(arXiv:2412.06264) 的技术解析：项目结构、三种范式（连续/离散/黎曼 Flow Matching）、概率路径与调度器、损失与求解器、流形与测地线实现，以及 2D/图像/文本示例、训练后调度器变换与 log 似然计算等使用指南。An Introduction to Flow Matching and Diffusion Modelshttps://steven-yl.github.io/mywebsite/an-introduction-to-flow-matching-and-diffusion-models/Sat, 28 Feb 2026 10:26:59 +0800Stevenhttps://github.com/steven-ylhttps://steven-yl.github.io/mywebsite/an-introduction-to-flow-matching-and-diffusion-models/<div class="featured-image"> <img src="/mywebsite/posts/images/an-introduction-to-flow-matching-and-diffusion-models.webp" referrerpolicy="no-referrer"> </div>《An Introduction to Flow Matching and Diffusion Models》全文技术解读：从生成即采样与 ODE/SDE 基础出发，系统介绍流模型与扩散模型、连续性方程与福克-普朗克方程、流匹配与得分匹配训练目标及其与 DDPM 的对应，并涵盖条件生成、无分类器引导（CFG）与 U-Net/DiT 等架构。Flow Matching Guide and Codehttps://steven-yl.github.io/mywebsite/flow-matching-guide-and-code/Sat, 28 Feb 2026 10:26:59 +0800Stevenhttps://github.com/steven-ylhttps://steven-yl.github.io/mywebsite/flow-matching-guide-and-code/<div class="featured-image"> <img src="/mywebsite/posts/images/flow-matching-guide-and-code.webp" referrerpolicy="no-referrer"> </div>《Flow Matching Guide and Code》全文技术解读：从流模型数学基础与欧氏空间 FM（概率路径、速度场、条件流匹配、线性/仿射条件流），到黎曼流形、离散 FM 与 Generator Matching 统一框架，并阐明与扩散模型、去噪分数匹配的关系。The Principles of Diffusion Modelshttps://steven-yl.github.io/mywebsite/the-principles-of-diffusion-models/Sat, 28 Feb 2026 10:26:59 +0800Stevenhttps://github.com/steven-ylhttps://steven-yl.github.io/mywebsite/the-principles-of-diffusion-models/<div class="featured-image"> <img src="/mywebsite/posts/images/the-principles-of-diffusion-models.webp" referrerpolicy="no-referrer"> </div>《The Principles of Diffusion Models》（arXiv:2510.21890）全文技术解读：从前向破坏过程与反向生成出发，系统梳理扩散模型的三种表述——变分视角（VAE→DDPM）、基于分数的视角（EBM→NCSN→分数 SDE）、基于流的视角（NF→流匹配），阐明条件化技巧与福克–普朗克方程下的统一；并涵盖引导生成、数值求解器、蒸馏与从零学习的流映射模型（CM/CTM/MF）等。Flow Matching Guide and Code 第5章解读：Non-Euclidean Flow Matchinghttps://steven-yl.github.io/mywebsite/flow-matching-guide-and-code-5-non-euclidean-flow-matching/Thu, 05 Mar 2026 00:00:00 +0000Stevenhttps://github.com/steven-ylhttps://steven-yl.github.io/mywebsite/flow-matching-guide-and-code-5-non-euclidean-flow-matching/第5章 Non-Euclidean Flow Matching 解读：从动机与黎曼流形设定出发，说明流形上的流、概率路径与连续性方程，边际化技巧（定理10）、RCFM 损失（定理11），以及测地线条件流与基于预度量的条件流；并对照欧氏 FM 与代码8（球面测地线 RCFM）做小结。