<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>Lemon&#39;s Blog</title>
    <link>https://coderlemon17.github.io/</link>
    <description>Recent content on Lemon&#39;s Blog</description>
    <image>
      <url>https://coderlemon17.github.io/%3Clink%20or%20path%20of%20image%20for%20opengraph,%20twitter-cards%3E</url>
      <link>https://coderlemon17.github.io/%3Clink%20or%20path%20of%20image%20for%20opengraph,%20twitter-cards%3E</link>
    </image>
    <generator>Hugo -- gohugo.io</generator>
    <lastBuildDate>Fri, 19 Aug 2022 15:08:49 +0800</lastBuildDate><atom:link href="https://coderlemon17.github.io/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Pytorch中的View, Reshape, Permute</title>
      <link>https://coderlemon17.github.io/posts/2022/08-19-view/</link>
      <pubDate>Fri, 19 Aug 2022 15:08:49 +0800</pubDate>
      
      <guid>https://coderlemon17.github.io/posts/2022/08-19-view/</guid>
      <description>Main takeaway 逻辑存储与内存存储 (Contiguous); Stride和Size View, Reshape 和 Permute ⭐: 本文是我最近写代码时的思考, 如果存在不合理或者更好的说明方式欢迎评论区提出. 一些小</description>
    </item>
    
    <item>
      <title>Self-labelling via simultaneous clustering and representation learning</title>
      <link>https://coderlemon17.github.io/posts/2022/07-18-sela/</link>
      <pubDate>Mon, 18 Jul 2022 15:08:49 +0800</pubDate>
      
      <guid>https://coderlemon17.github.io/posts/2022/07-18-sela/</guid>
      <description>论文信息: ICLR 2020 (Spotlight); 目前引用量: 329 论文链接: https://arxiv.org/pdf/1911.05371.pdf 0. Main Takeaway 通过加入equipartition constraint, 本文将simultaneous clustering and representation learning问题</description>
    </item>
    
    <item>
      <title>Optimal Transport入门简述</title>
      <link>https://coderlemon17.github.io/posts/2022/07-16-ot/</link>
      <pubDate>Sat, 16 Jul 2022 14:26:33 +0800</pubDate>
      
      <guid>https://coderlemon17.github.io/posts/2022/07-16-ot/</guid>
      <description>0. Main Takeaway 本文主要是对Notes on Optimal Transport 的翻译及整理, 同时重构了原文提供的代码, 整理到了一个Jupyter Notebook中并增加了注释, 方便理解</description>
    </item>
    
    <item>
      <title>详解Markov Chain Monte Carlo (MCMC): 从拒绝-接受采样到Gibbs Sampling</title>
      <link>https://coderlemon17.github.io/posts/2022/05-11-mcmc/</link>
      <pubDate>Wed, 11 May 2022 20:30:33 +0800</pubDate>
      
      <guid>https://coderlemon17.github.io/posts/2022/05-11-mcmc/</guid>
      <description>0. Main Takeaway Sec 1. 介绍了蒙特卡洛采样 (MC) 和拒绝-接受采样. Sec 2. 介绍了马尔可夫链(MC)和基于马尔可夫链的采样. Sec 3. 介绍了马尔可夫链蒙特卡洛采样(MCM</description>
    </item>
    
    <item>
      <title>torch.nn.parallel.DistributedDataParallel: 快速上手</title>
      <link>https://coderlemon17.github.io/posts/2022/02-14-ddp/</link>
      <pubDate>Mon, 14 Feb 2022 21:32:33 +0800</pubDate>
      
      <guid>https://coderlemon17.github.io/posts/2022/02-14-ddp/</guid>
      <description>0. 前言 Node: 一个节点, 可以理解为一台电脑. Device: 工作设备, 可以简单理解为一张卡, 即一个GPU. Process: 一个进程, 可以简单理解为一个Python程序. Threading: 一个</description>
    </item>
    
    
  </channel>
</rss>
