mirror of
https://github.com/labmlai/annotated_deep_learning_paper_implementations.git
synced 2025-08-16 02:41:38 +08:00
726 lines
44 KiB
HTML
726 lines
44 KiB
HTML
<!DOCTYPE html>
|
||
<html lang="ja">
|
||
<head>
|
||
<meta http-equiv="content-type" content="text/html;charset=utf-8"/>
|
||
<meta name="viewport" content="width=device-width, initial-scale=1.0"/>
|
||
<meta name="description" content="「必要なのはパッチだけ？」という論文のPyTorch実装/チュートリアル"/>
|
||
|
||
<meta name="twitter:card" content="summary"/>
|
||
<meta name="twitter:image:src" content="https://avatars1.githubusercontent.com/u/64068543?s=400&v=4"/>
|
||
<meta name="twitter:title" content="必要なのはパッチだけ？(コンバージョンミキサー)"/>
|
||
<meta name="twitter:description" content="「必要なのはパッチだけ？」という論文のPyTorch実装/チュートリアル"/>
|
||
<meta name="twitter:site" content="@labmlai"/>
|
||
<meta name="twitter:creator" content="@labmlai"/>
|
||
|
||
<meta property="og:url" content="https://nn.labml.ai/conv_mixer/index.html"/>
|
||
<meta property="og:title" content="必要なのはパッチだけ？(コンバージョンミキサー)"/>
|
||
<meta property="og:image" content="https://avatars1.githubusercontent.com/u/64068543?s=400&v=4"/>
|
||
<meta property="og:site_name" content="必要なのはパッチだけ？(コンバージョンミキサー)"/>
|
||
<meta property="og:type" content="object"/>
|
||
<meta property="og:title" content="必要なのはパッチだけ？(コンバージョンミキサー)"/>
|
||
<meta property="og:description" content="「必要なのはパッチだけ？」という論文のPyTorch実装/チュートリアル"/>
|
||
|
||
<title>必要なのはパッチだけ？(コンバージョンミキサー)</title>
|
||
<link rel="shortcut icon" href="/icon.png"/>
|
||
<link rel="stylesheet" href="../pylit.css?v=1">
|
||
<link rel="canonical" href="https://nn.labml.ai/conv_mixer/index.html"/>
|
||
<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/katex@0.13.18/dist/katex.min.css" integrity="sha384-zTROYFVGOfTw7JV7KUu8udsvW2fx4lWOsCEDqhBreBwlHI4ioVRtmIvEThzJHGET" crossorigin="anonymous">
|
||
|
||
<!-- Global site tag (gtag.js) - Google Analytics -->
|
||
<script async src="https://www.googletagmanager.com/gtag/js?id=G-4V3HC8HBLH"></script>
|
||
<script>
|
||
window.dataLayer = window.dataLayer || [];
|
||
|
||
function gtag() {
|
||
dataLayer.push(arguments);
|
||
}
|
||
|
||
gtag('js', new Date());
|
||
|
||
gtag('config', 'G-4V3HC8HBLH');
|
||
</script>
|
||
</head>
|
||
<body>
|
||
<div id='container'>
|
||
<div id="background"></div>
|
||
<div class='section'>
|
||
<div class='docs'>
|
||
<p>
|
||
<a class="parent" href="/">home</a>
|
||
<a class="parent" href="index.html">conv_mixer</a>
|
||
</p>
|
||
<p>
|
||
<a href="https://github.com/labmlai/annotated_deep_learning_paper_implementations" target="_blank">
|
||
<img alt="Github"
|
||
src="https://img.shields.io/github/stars/labmlai/annotated_deep_learning_paper_implementations?style=social"
|
||
style="max-width:100%;"/></a>
|
||
<a href="https://twitter.com/labmlai" rel="nofollow" target="_blank">
|
||
<img alt="Twitter"
|
||
src="https://img.shields.io/twitter/follow/labmlai?style=social"
|
||
style="max-width:100%;"/></a>
|
||
</p>
|
||
<p>
|
||
<a href="https://github.com/labmlai/annotated_deep_learning_paper_implementations/tree/master/labml_nn/conv_mixer/__init__.py" target="_blank">
|
||
View code on Github</a>
|
||
</p>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-0'>
|
||
<div class='docs doc-strings'>
|
||
<div class='section-link'>
|
||
<a href='#section-0'>#</a>
|
||
</div>
|
||
<h1>必要なのはパッチだけ？(コンバージョンミキサー</h1>)
|
||
<p><a href="https://pytorch.org">これは紙のパッチをPyTorchで実装したものです</a><a href="https://papers.labml.ai/paper/2201.09792">。必要なのはパッチだけですか</a>？</p>。
|
||
<p><img alt="ConvMixer diagram from the paper" src="conv_mixer.png"></p>
|
||
<p><a href="../transformers/mlp_mixer/index.html">ConvMixerはMLPミキサーに似ています。</a></p><a href="../transformers/feed_forward.html">MLP-Mixerは、空間次元全体にMLPを適用し、次にチャネル次元全体にMLPを適用することで、空間次元とチャネル次元の混合を分離します（空間MLPはvITのアテンションに代わり、<a href="../transformers/vit/index.html">チャネルMLPはVITのFFNです</a>）。</a>
|
||
<p>ConvMixerは、<span ><span class="katex"><span aria-hidden="true" class="katex-html"><span class="base"><span class="strut" style="height:0.72777em;vertical-align:-0.08333em;"></span><span class="mord coloredeq eqa" style=""><span class="mord" style="">1</span><span class="mspace" style="margin-right:0.2222222222222222em;"></span><span class="mbin" style="">×</span><span class="mspace" style="margin-right:0.2222222222222222em;"></span><span class="mord" style="">1</span></span></span></span></span></span>チャンネルミキシングにはコンボリューションを使用し、空間ミキシングには奥行きコンボリューションを使用します。スペース全体でフルMLPではなく畳み込みなので、VITやMLPミキサーとは対照的に、近くのバッチのみをミキシングします。また、MLPミキサーはミキシングごとに2層のMLPを使用し、ConvMixerはミキシングごとに1層のMLPを使用します</p>。
|
||
<p>この論文では、チャネルミキシング全体の残留接続を削除し（点単位の畳み込み）、空間ミキシングでは残留接続のみにする（深さ方向の畳み込み）ことを推奨しています。また、</p><a href="../normalization/batch_norm/index.html"><a href="../normalization/layer_norm/index.html">レイヤー正規化の代わりにバッチ正規化を使用します</a></a>。
|
||
<p>これは<a href="experiment.html">、CIFAR-10 で ConvMixer をトレーニングする実験です</a>。</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">36</span><span></span><span class="kn">import</span> <span class="nn">torch</span>
|
||
<span class="lineno">37</span><span class="kn">from</span> <span class="nn">torch</span> <span class="kn">import</span> <span class="n">nn</span>
|
||
<span class="lineno">38</span>
|
||
<span class="lineno">39</span><span class="kn">from</span> <span class="nn">labml_helpers.module</span> <span class="kn">import</span> <span class="n">Module</span>
|
||
<span class="lineno">40</span><span class="kn">from</span> <span class="nn">labml_nn.utils</span> <span class="kn">import</span> <span class="n">clone_module_list</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-1'>
|
||
<div class='docs doc-strings'>
|
||
<div class='section-link'>
|
||
<a href='#section-1'>#</a>
|
||
</div>
|
||
<p><a id="ConvMixerLayer"></a></p>
|
||
<h2>ConvMixer レイヤー</h2>
|
||
<p>これは単一の ConvMixer レイヤーです。モデルにはこれらのシリーズがあります。</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">43</span><span class="k">class</span> <span class="nc">ConvMixerLayer</span><span class="p">(</span><span class="n">Module</span><span class="p">):</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-2'>
|
||
<div class='docs doc-strings'>
|
||
<div class='section-link'>
|
||
<a href='#section-2'>#</a>
|
||
</div>
|
||
<ul><li><code class="highlight"><span></span><span class="n">d_model</span></code>
|
||
はパッチ埋め込みのチャンネル数、<span ><span class="katex"><span aria-hidden="true" class="katex-html"><span class="base"><span class="strut" style="height:0.69444em;vertical-align:0em;"></span><span class="mord coloredeq eqd" style=""><span class="mord mathnormal" style="">h</span></span></span></span></span></span></li>
|
||
<li><code class="highlight"><span></span><span class="n">kernel_size</span></code>
|
||
は空間畳み込みのカーネルの大きさです <span ><span class="katex"><span aria-hidden="true" class="katex-html"><span class="base"><span class="strut" style="height:0.69444em;vertical-align:0em;"></span><span class="mord mathnormal" style="margin-right:0.03148em;">k</span></span></span></span></span></li></ul>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">52</span> <span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">d_model</span><span class="p">:</span> <span class="nb">int</span><span class="p">,</span> <span class="n">kernel_size</span><span class="p">:</span> <span class="nb">int</span><span class="p">):</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-3'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-3'>#</a>
|
||
</div>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">57</span> <span class="nb">super</span><span class="p">()</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-4'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-4'>#</a>
|
||
</div>
|
||
<p>深度方向の畳み込みは、チャンネルごとに別々の畳み込みになります。これは、グループの数がチャネル数と等しい畳み込み層で行います。そのため、各チャンネルはそれぞれ独自のグループになります。</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">61</span> <span class="bp">self</span><span class="o">.</span><span class="n">depth_wise_conv</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="n">d_model</span><span class="p">,</span> <span class="n">d_model</span><span class="p">,</span>
|
||
<span class="lineno">62</span> <span class="n">kernel_size</span><span class="o">=</span><span class="n">kernel_size</span><span class="p">,</span>
|
||
<span class="lineno">63</span> <span class="n">groups</span><span class="o">=</span><span class="n">d_model</span><span class="p">,</span>
|
||
<span class="lineno">64</span> <span class="n">padding</span><span class="o">=</span><span class="p">(</span><span class="n">kernel_size</span> <span class="o">-</span> <span class="mi">1</span><span class="p">)</span> <span class="o">//</span> <span class="mi">2</span><span class="p">)</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-5'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-5'>#</a>
|
||
</div>
|
||
<p>深さ方向の畳み込み後のアクティベーション</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">66</span> <span class="bp">self</span><span class="o">.</span><span class="n">act1</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">GELU</span><span class="p">()</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-6'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-6'>#</a>
|
||
</div>
|
||
<p>深さ方向の畳み込み後の正規化</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">68</span> <span class="bp">self</span><span class="o">.</span><span class="n">norm1</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">BatchNorm2d</span><span class="p">(</span><span class="n">d_model</span><span class="p">)</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-7'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-7'>#</a>
|
||
</div>
|
||
<p>点単位の畳み込みは畳み込みです。つまり、<span ><span class="katex"><span aria-hidden="true" class="katex-html"><span class="base"><span class="strut" style="height:0.72777em;vertical-align:-0.08333em;"></span><span class="mord coloredeq eqa" style=""><span class="mord" style="">1</span><span class="mspace" style="margin-right:0.2222222222222222em;"></span><span class="mbin" style="">×</span><span class="mspace" style="margin-right:0.2222222222222222em;"></span><span class="mord" style="">1</span></span></span></span></span></span>パッチ埋め込みの線形変換です</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">72</span> <span class="bp">self</span><span class="o">.</span><span class="n">point_wise_conv</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="n">d_model</span><span class="p">,</span> <span class="n">d_model</span><span class="p">,</span> <span class="n">kernel_size</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-8'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-8'>#</a>
|
||
</div>
|
||
<p>点単位の畳み込み後のアクティベーション</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">74</span> <span class="bp">self</span><span class="o">.</span><span class="n">act2</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">GELU</span><span class="p">()</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-9'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-9'>#</a>
|
||
</div>
|
||
<p>点単位の畳み込み後の正規化</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">76</span> <span class="bp">self</span><span class="o">.</span><span class="n">norm2</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">BatchNorm2d</span><span class="p">(</span><span class="n">d_model</span><span class="p">)</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-10'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-10'>#</a>
|
||
</div>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">78</span> <span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">x</span><span class="p">:</span> <span class="n">torch</span><span class="o">.</span><span class="n">Tensor</span><span class="p">):</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-11'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-11'>#</a>
|
||
</div>
|
||
<p>深さ方向の畳み込みの周りの残差結合について</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">80</span> <span class="n">residual</span> <span class="o">=</span> <span class="n">x</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-12'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-12'>#</a>
|
||
</div>
|
||
<p>深度方向の畳み込み、活性化、正規化</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">83</span> <span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">depth_wise_conv</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
|
||
<span class="lineno">84</span> <span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">act1</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
|
||
<span class="lineno">85</span> <span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">norm1</span><span class="p">(</span><span class="n">x</span><span class="p">)</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-13'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-13'>#</a>
|
||
</div>
|
||
<p>残余接続を追加</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">88</span> <span class="n">x</span> <span class="o">+=</span> <span class="n">residual</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-14'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-14'>#</a>
|
||
</div>
|
||
<p>点単位の畳み込み、活性化、正規化</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">91</span> <span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">point_wise_conv</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
|
||
<span class="lineno">92</span> <span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">act2</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
|
||
<span class="lineno">93</span> <span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">norm2</span><span class="p">(</span><span class="n">x</span><span class="p">)</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-15'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-15'>#</a>
|
||
</div>
|
||
<p></p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">96</span> <span class="k">return</span> <span class="n">x</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-16'>
|
||
<div class='docs doc-strings'>
|
||
<div class='section-link'>
|
||
<a href='#section-16'>#</a>
|
||
</div>
|
||
<p><a id="PatchEmbeddings"></a></p>
|
||
<h2>パッチ埋め込みを入手</h2>
|
||
<p>これにより、<span ><span class="katex"><span aria-hidden="true" class="katex-html"><span class="base"><span class="strut" style="height:0.7777700000000001em;vertical-align:-0.19444em;"></span><span class="mord coloredeq eqf" style=""><span class="mord mathnormal" style="">p</span></span><span class="mspace" style="margin-right:0.2222222222222222em;"></span><span class="mbin">×</span><span class="mspace" style="margin-right:0.2222222222222222em;"></span></span><span class="base"><span class="strut" style="height:0.625em;vertical-align:-0.19444em;"></span><span class="mord coloredeq eqf" style=""><span class="mord mathnormal" style="">p</span></span></span></span></span></span>画像が複数のサイズのパッチに分割され、各パッチが埋め込まれます。</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">99</span><span class="k">class</span> <span class="nc">PatchEmbeddings</span><span class="p">(</span><span class="n">Module</span><span class="p">):</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-17'>
|
||
<div class='docs doc-strings'>
|
||
<div class='section-link'>
|
||
<a href='#section-17'>#</a>
|
||
</div>
|
||
<ul><li><code class="highlight"><span></span><span class="n">d_model</span></code>
|
||
パッチ埋め込みのチャンネル数です <span ><span class="katex"><span aria-hidden="true" class="katex-html"><span class="base"><span class="strut" style="height:0.69444em;vertical-align:0em;"></span><span class="mord coloredeq eqd" style=""><span class="mord mathnormal" style="">h</span></span></span></span></span></span></li>
|
||
<li><code class="highlight"><span></span><span class="n">patch_size</span></code>
|
||
はパッチのサイズ、<span ><span class="katex"><span aria-hidden="true" class="katex-html"><span class="base"><span class="strut" style="height:0.625em;vertical-align:-0.19444em;"></span><span class="mord coloredeq eqf" style=""><span class="mord mathnormal" style="">p</span></span></span></span></span></span></li>
|
||
<li><code class="highlight"><span></span><span class="n">in_channels</span></code>
|
||
は入力画像のチャンネル数 (RGB の場合は 3)</li></ul>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">108</span> <span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">d_model</span><span class="p">:</span> <span class="nb">int</span><span class="p">,</span> <span class="n">patch_size</span><span class="p">:</span> <span class="nb">int</span><span class="p">,</span> <span class="n">in_channels</span><span class="p">:</span> <span class="nb">int</span><span class="p">):</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-18'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-18'>#</a>
|
||
</div>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">114</span> <span class="nb">super</span><span class="p">()</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-19'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-19'>#</a>
|
||
</div>
|
||
<p>カーネルサイズでストライドの長さがパッチサイズと同じコンボリューションレイヤーを作成します。これは、画像をパッチに分割し、各パッチで線形変換を行うのと同じです</p>。
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">119</span> <span class="bp">self</span><span class="o">.</span><span class="n">conv</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="n">in_channels</span><span class="p">,</span> <span class="n">d_model</span><span class="p">,</span> <span class="n">kernel_size</span><span class="o">=</span><span class="n">patch_size</span><span class="p">,</span> <span class="n">stride</span><span class="o">=</span><span class="n">patch_size</span><span class="p">)</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-20'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-20'>#</a>
|
||
</div>
|
||
<p>アクティベーション機能</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">121</span> <span class="bp">self</span><span class="o">.</span><span class="n">act</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">GELU</span><span class="p">()</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-21'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-21'>#</a>
|
||
</div>
|
||
<p>バッチ正規化</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">123</span> <span class="bp">self</span><span class="o">.</span><span class="n">norm</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">BatchNorm2d</span><span class="p">(</span><span class="n">d_model</span><span class="p">)</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-22'>
|
||
<div class='docs doc-strings'>
|
||
<div class='section-link'>
|
||
<a href='#section-22'>#</a>
|
||
</div>
|
||
<ul><li><code class="highlight"><span></span><span class="n">x</span></code>
|
||
形状の入力画像です <code class="highlight"><span></span><span class="p">[</span><span class="n">batch_size</span><span class="p">,</span> <span class="n">channels</span><span class="p">,</span> <span class="n">height</span><span class="p">,</span> <span class="n">width</span><span class="p">]</span></code>
|
||
</li></ul>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">125</span> <span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">x</span><span class="p">:</span> <span class="n">torch</span><span class="o">.</span><span class="n">Tensor</span><span class="p">):</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-23'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-23'>#</a>
|
||
</div>
|
||
<p>畳み込み層を適用</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">130</span> <span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">conv</span><span class="p">(</span><span class="n">x</span><span class="p">)</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-24'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-24'>#</a>
|
||
</div>
|
||
<p>アクティベーションと正規化</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">132</span> <span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">act</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
|
||
<span class="lineno">133</span> <span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">norm</span><span class="p">(</span><span class="n">x</span><span class="p">)</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-25'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-25'>#</a>
|
||
</div>
|
||
<p></p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">136</span> <span class="k">return</span> <span class="n">x</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-26'>
|
||
<div class='docs doc-strings'>
|
||
<div class='section-link'>
|
||
<a href='#section-26'>#</a>
|
||
</div>
|
||
<p><a id="ClassificationHead"></a></p>
|
||
<h2>分類責任者</h2>
|
||
<p>平均プーリング（すべてのパッチ埋め込みの平均を取る）と最終的な線形変換を行って、画像クラスの対数確率を予測します。</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">139</span><span class="k">class</span> <span class="nc">ClassificationHead</span><span class="p">(</span><span class="n">Module</span><span class="p">):</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-27'>
|
||
<div class='docs doc-strings'>
|
||
<div class='section-link'>
|
||
<a href='#section-27'>#</a>
|
||
</div>
|
||
<ul><li><code class="highlight"><span></span><span class="n">d_model</span></code>
|
||
はパッチ埋め込みのチャンネル数、<span ><span class="katex"><span aria-hidden="true" class="katex-html"><span class="base"><span class="strut" style="height:0.69444em;vertical-align:0em;"></span><span class="mord coloredeq eqd" style=""><span class="mord mathnormal" style="">h</span></span></span></span></span></span></li>
|
||
<li><code class="highlight"><span></span><span class="n">n_classes</span></code>
|
||
分類タスク内のクラス数です</li></ul>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">149</span> <span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">d_model</span><span class="p">:</span> <span class="nb">int</span><span class="p">,</span> <span class="n">n_classes</span><span class="p">:</span> <span class="nb">int</span><span class="p">):</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-28'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-28'>#</a>
|
||
</div>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">154</span> <span class="nb">super</span><span class="p">()</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-29'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-29'>#</a>
|
||
</div>
|
||
<p>アベレージプール</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">156</span> <span class="bp">self</span><span class="o">.</span><span class="n">pool</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">AdaptiveAvgPool2d</span><span class="p">((</span><span class="mi">1</span><span class="p">,</span> <span class="mi">1</span><span class="p">))</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-30'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-30'>#</a>
|
||
</div>
|
||
<p>リニアレイヤー</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">158</span> <span class="bp">self</span><span class="o">.</span><span class="n">linear</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="n">d_model</span><span class="p">,</span> <span class="n">n_classes</span><span class="p">)</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-31'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-31'>#</a>
|
||
</div>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">160</span> <span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">x</span><span class="p">:</span> <span class="n">torch</span><span class="o">.</span><span class="n">Tensor</span><span class="p">):</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-32'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-32'>#</a>
|
||
</div>
|
||
<p>平均プーリング</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">162</span> <span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">pool</span><span class="p">(</span><span class="n">x</span><span class="p">)</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-33'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-33'>#</a>
|
||
</div>
|
||
<p><code class="highlight"><span></span><span class="n">x</span></code>
|
||
埋め込みを入れると、形が整います <code class="highlight"><span></span><span class="p">[</span><span class="n">batch_size</span><span class="p">,</span> <span class="n">d_model</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="mi">1</span><span class="p">]</span></code>
|
||
</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">164</span> <span class="n">x</span> <span class="o">=</span> <span class="n">x</span><span class="p">[:,</span> <span class="p">:,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">0</span><span class="p">]</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-34'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-34'>#</a>
|
||
</div>
|
||
<p>リニアレイヤー</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">166</span> <span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">linear</span><span class="p">(</span><span class="n">x</span><span class="p">)</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-35'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-35'>#</a>
|
||
</div>
|
||
<p></p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">169</span> <span class="k">return</span> <span class="n">x</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-36'>
|
||
<div class='docs doc-strings'>
|
||
<div class='section-link'>
|
||
<a href='#section-36'>#</a>
|
||
</div>
|
||
<h2>コンバージョンミキサー</h2>
|
||
<p>これにより、パッチ埋め込みブロック、多数の ConvMixer レイヤー、および分類ヘッドが組み合わされます。</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">172</span><span class="k">class</span> <span class="nc">ConvMixer</span><span class="p">(</span><span class="n">Module</span><span class="p">):</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-37'>
|
||
<div class='docs doc-strings'>
|
||
<div class='section-link'>
|
||
<a href='#section-37'>#</a>
|
||
</div>
|
||
<ul><li><code class="highlight"><span></span><span class="n">conv_mixer_layer</span></code>
|
||
単一の <a href="#ConvMixerLayer">ConvMixer</a> レイヤーのコピーです。そのコピーを作成して ConvMixer</li> を作成します。<code class="highlight"><span></span><span class="n">n_layers</span></code>
|
||
|
||
<li><code class="highlight"><span></span><span class="n">n_layers</span></code>
|
||
は ConvMixer レイヤーの数 (または深さ) です。<span ><span class="katex"><span aria-hidden="true" class="katex-html"><span class="base"><span class="strut" style="height:0.69444em;vertical-align:0em;"></span><span class="mord mathnormal">d</span></span></span></span></span></li>
|
||
<li><code class="highlight"><span></span><span class="n">patch_emb</span></code>
|
||
<a href="#PatchEmbeddings">パッチ埋め込みレイヤーです</a>。</li>
|
||
<li><code class="highlight"><span></span><span class="n">classification</span></code>
|
||
<a href="#ClassificationHead">分類責任者です</a>。</li></ul>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">179</span> <span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">conv_mixer_layer</span><span class="p">:</span> <span class="n">ConvMixerLayer</span><span class="p">,</span> <span class="n">n_layers</span><span class="p">:</span> <span class="nb">int</span><span class="p">,</span>
|
||
<span class="lineno">180</span> <span class="n">patch_emb</span><span class="p">:</span> <span class="n">PatchEmbeddings</span><span class="p">,</span>
|
||
<span class="lineno">181</span> <span class="n">classification</span><span class="p">:</span> <span class="n">ClassificationHead</span><span class="p">):</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-38'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-38'>#</a>
|
||
</div>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">189</span> <span class="nb">super</span><span class="p">()</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-39'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-39'>#</a>
|
||
</div>
|
||
<p>パッチ埋め込み</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">191</span> <span class="bp">self</span><span class="o">.</span><span class="n">patch_emb</span> <span class="o">=</span> <span class="n">patch_emb</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-40'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-40'>#</a>
|
||
</div>
|
||
<p>分類ヘッド</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">193</span> <span class="bp">self</span><span class="o">.</span><span class="n">classification</span> <span class="o">=</span> <span class="n">classification</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-41'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-41'>#</a>
|
||
</div>
|
||
<p><a href="#ConvMixerLayer">ConvMixer</a> レイヤーのコピーを作成</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">195</span> <span class="bp">self</span><span class="o">.</span><span class="n">conv_mixer_layers</span> <span class="o">=</span> <span class="n">clone_module_list</span><span class="p">(</span><span class="n">conv_mixer_layer</span><span class="p">,</span> <span class="n">n_layers</span><span class="p">)</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-42'>
|
||
<div class='docs doc-strings'>
|
||
<div class='section-link'>
|
||
<a href='#section-42'>#</a>
|
||
</div>
|
||
<ul><li><code class="highlight"><span></span><span class="n">x</span></code>
|
||
形状の入力画像です <code class="highlight"><span></span><span class="p">[</span><span class="n">batch_size</span><span class="p">,</span> <span class="n">channels</span><span class="p">,</span> <span class="n">height</span><span class="p">,</span> <span class="n">width</span><span class="p">]</span></code>
|
||
</li></ul>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">197</span> <span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">x</span><span class="p">:</span> <span class="n">torch</span><span class="o">.</span><span class="n">Tensor</span><span class="p">):</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-43'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-43'>#</a>
|
||
</div>
|
||
<p>パッチの埋め込みを入手してください。<code class="highlight"><span></span><span class="p">[</span><span class="n">batch_size</span><span class="p">,</span> <span class="n">d_model</span><span class="p">,</span> <span class="n">height</span> <span class="o">/</span> <span class="n">patch_size</span><span class="p">,</span> <span class="n">width</span> <span class="o">/</span> <span class="n">patch_size</span><span class="p">]</span></code>
|
||
これにより形状のテンソルが得られます</p>。
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">202</span> <span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">patch_emb</span><span class="p">(</span><span class="n">x</span><span class="p">)</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-44'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-44'>#</a>
|
||
</div>
|
||
<p><a href="#ConvMixerLayer">ConvMixer レイヤーをパススルーする</a></p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">205</span> <span class="k">for</span> <span class="n">layer</span> <span class="ow">in</span> <span class="bp">self</span><span class="o">.</span><span class="n">conv_mixer_layers</span><span class="p">:</span>
|
||
<span class="lineno">206</span> <span class="n">x</span> <span class="o">=</span> <span class="n">layer</span><span class="p">(</span><span class="n">x</span><span class="p">)</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-45'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-45'>#</a>
|
||
</div>
|
||
<p>ロジットを取得するための分類ヘッド</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">209</span> <span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">classification</span><span class="p">(</span><span class="n">x</span><span class="p">)</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='section' id='section-46'>
|
||
<div class='docs'>
|
||
<div class='section-link'>
|
||
<a href='#section-46'>#</a>
|
||
</div>
|
||
<p></p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">212</span> <span class="k">return</span> <span class="n">x</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='footer'>
|
||
<a href="https://papers.labml.ai">Trending Research Papers</a>
|
||
<a href="https://labml.ai">labml.ai</a>
|
||
</div>
|
||
</div>
|
||
<script src=../interactive.js?v=1"></script>
|
||
<script>
|
||
function handleImages() {
|
||
var images = document.querySelectorAll('p>img')
|
||
|
||
for (var i = 0; i < images.length; ++i) {
|
||
handleImage(images[i])
|
||
}
|
||
}
|
||
|
||
function handleImage(img) {
|
||
img.parentElement.style.textAlign = 'center'
|
||
|
||
var modal = document.createElement('div')
|
||
modal.id = 'modal'
|
||
|
||
var modalContent = document.createElement('div')
|
||
modal.appendChild(modalContent)
|
||
|
||
var modalImage = document.createElement('img')
|
||
modalContent.appendChild(modalImage)
|
||
|
||
var span = document.createElement('span')
|
||
span.classList.add('close')
|
||
span.textContent = 'x'
|
||
modal.appendChild(span)
|
||
|
||
img.onclick = function () {
|
||
console.log('clicked')
|
||
document.body.appendChild(modal)
|
||
modalImage.src = img.src
|
||
}
|
||
|
||
span.onclick = function () {
|
||
document.body.removeChild(modal)
|
||
}
|
||
}
|
||
|
||
handleImages()
|
||
</script>
|
||
</body>
|
||
</html> |