mirror of
https://github.com/labmlai/annotated_deep_learning_paper_implementations.git
synced 2025-08-17 11:20:32 +08:00
693 lines
39 KiB
HTML
693 lines
39 KiB
HTML
<!DOCTYPE html>
|
|
<html>
|
|
<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="A PyTorch implementation/tutorial of the paper "An Image Is Worth 16x16 Words: Transformers For Image Recognition At Scale""/>
|
|
|
|
<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="Vision Transformer (ViT)"/>
|
|
<meta name="twitter:description" content="A PyTorch implementation/tutorial of the paper "An Image Is Worth 16x16 Words: Transformers For Image Recognition At Scale""/>
|
|
<meta name="twitter:site" content="@labmlai"/>
|
|
<meta name="twitter:creator" content="@labmlai"/>
|
|
|
|
<meta property="og:url" content="https://nn.labml.ai/transformers/vit/index.html"/>
|
|
<meta property="og:title" content="Vision Transformer (ViT)"/>
|
|
<meta property="og:image" content="https://avatars1.githubusercontent.com/u/64068543?s=400&v=4"/>
|
|
<meta property="og:site_name" content="LabML Neural Networks"/>
|
|
<meta property="og:type" content="object"/>
|
|
<meta property="og:title" content="Vision Transformer (ViT)"/>
|
|
<meta property="og:description" content="A PyTorch implementation/tutorial of the paper "An Image Is Worth 16x16 Words: Transformers For Image Recognition At Scale""/>
|
|
|
|
<title>Vision Transformer (ViT)</title>
|
|
<link rel="shortcut icon" href="/icon.png"/>
|
|
<link rel="stylesheet" href="../../pylit.css">
|
|
<link rel="canonical" href="https://nn.labml.ai/transformers/vit/index.html"/>
|
|
<!-- 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">transformers</a>
|
|
<a class="parent" href="index.html">vit</a>
|
|
</p>
|
|
<p>
|
|
|
|
<a href="https://github.com/labmlai/annotated_deep_learning_paper_implementations/tree/master/labml_nn/transformers/vit/__init__.py">
|
|
<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">
|
|
<img alt="Twitter"
|
|
src="https://img.shields.io/twitter/follow/labmlai?style=social"
|
|
style="max-width:100%;"/></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>Vision Transformer (ViT)</h1>
|
|
<p>This is a <a href="https://pytorch.org">PyTorch</a> implementation of the paper
|
|
<a href="https://papers.labml.ai/paper/2010.11929">An Image Is Worth 16x16 Words: Transformers For Image Recognition At Scale</a>.</p>
|
|
<p>Vision transformer applies a pure transformer to images
|
|
without any convolution layers.
|
|
They split the image into patches and apply a transformer on patch embeddings.
|
|
<a href="#PathEmbeddings">Patch embeddings</a> are generated by applying a simple linear transformation
|
|
to the flattened pixel values of the patch.
|
|
Then a standard transformer encoder is fed with the patch embeddings, along with a
|
|
classification token <code>[CLS]</code>.
|
|
The encoding on the <code>[CLS]</code> token is used to classify the image with an MLP.</p>
|
|
<p>When feeding the transformer with the patches, learned positional embeddings are
|
|
added to the patch embeddings, because the patch embeddings do not have any information
|
|
about where that patch is from.
|
|
The positional embeddings are a set of vectors for each patch location that get trained
|
|
with gradient descent along with other parameters.</p>
|
|
<p>ViTs perform well when they are pre-trained on large datasets.
|
|
The paper suggests pre-training them with an MLP classification head and
|
|
then using a single linear layer when fine-tuning.
|
|
The paper beats SOTA with a ViT pre-trained on a 300 million image dataset.
|
|
They also use higher resolution images during inference while keeping the
|
|
patch size the same.
|
|
The positional embeddings for new patch locations are calculated by interpolating
|
|
learning positional embeddings.</p>
|
|
<p>Here’s <a href="experiment.html">an experiment</a> that trains ViT on CIFAR-10.
|
|
This doesn’t do very well because it’s trained on a small dataset.
|
|
It’s a simple experiment that anyone can run and play with ViTs.</p>
|
|
<p><a href="https://app.labml.ai/run/8b531d9ce3dc11eb84fc87df6756eb8f"><img alt="View Run" src="https://img.shields.io/badge/labml-experiment-brightgreen" /></a></p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">45</span><span></span><span class="kn">import</span> <span class="nn">torch</span>
|
|
<span class="lineno">46</span><span class="kn">from</span> <span class="nn">torch</span> <span class="kn">import</span> <span class="n">nn</span>
|
|
<span class="lineno">47</span>
|
|
<span class="lineno">48</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">49</span><span class="kn">from</span> <span class="nn">labml_nn.transformers</span> <span class="kn">import</span> <span class="n">TransformerLayer</span>
|
|
<span class="lineno">50</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="PatchEmbeddings"></p>
|
|
<h2>Get patch embeddings</h2>
|
|
<p></a></p>
|
|
<p>The paper splits the image into patches of equal size and do a linear transformation
|
|
on the flattened pixels for each patch.</p>
|
|
<p>We implement the same thing through a convolution layer, because it’s simpler to implement.</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">53</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-2'>
|
|
<div class='docs doc-strings'>
|
|
<div class='section-link'>
|
|
<a href='#section-2'>#</a>
|
|
</div>
|
|
<ul>
|
|
<li><code>d_model</code> is the transformer embeddings size</li>
|
|
<li><code>patch_size</code> is the size of the patch</li>
|
|
<li><code>in_channels</code> is the number of channels in the input image (3 for rgb)</li>
|
|
</ul>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">65</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-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">71</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>We create a convolution layer with a kernel size and and stride length equal to patch size.
|
|
This is equivalent to splitting the image into patches and doing a linear
|
|
transformation on each patch.</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">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">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-5'>
|
|
<div class='docs doc-strings'>
|
|
<div class='section-link'>
|
|
<a href='#section-5'>#</a>
|
|
</div>
|
|
<ul>
|
|
<li><code>x</code> is the input image of shape <code>[batch_size, channels, height, width]</code></li>
|
|
</ul>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">78</span> <span class="k">def</span> <span class="fm">__call__</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-6'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-6'>#</a>
|
|
</div>
|
|
<p>Apply convolution layer</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">conv</span><span class="p">(</span><span class="n">x</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>Get the shape.</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">85</span> <span class="n">bs</span><span class="p">,</span> <span class="n">c</span><span class="p">,</span> <span class="n">h</span><span class="p">,</span> <span class="n">w</span> <span class="o">=</span> <span class="n">x</span><span class="o">.</span><span class="n">shape</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>Rearrange to shape <code>[patches, batch_size, d_model]</code></p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">87</span> <span class="n">x</span> <span class="o">=</span> <span class="n">x</span><span class="o">.</span><span class="n">permute</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">1</span><span class="p">)</span>
|
|
<span class="lineno">88</span> <span class="n">x</span> <span class="o">=</span> <span class="n">x</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="n">h</span> <span class="o">*</span> <span class="n">w</span><span class="p">,</span> <span class="n">bs</span><span class="p">,</span> <span class="n">c</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>Return the patch embeddings</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">91</span> <span class="k">return</span> <span class="n">x</span></pre></div>
|
|
</div>
|
|
</div>
|
|
<div class='section' id='section-10'>
|
|
<div class='docs doc-strings'>
|
|
<div class='section-link'>
|
|
<a href='#section-10'>#</a>
|
|
</div>
|
|
<p><a id="LearnedPositionalEmbeddings"></p>
|
|
<h2>Add parameterized positional encodings</h2>
|
|
<p></a></p>
|
|
<p>This adds learned positional embeddings to patch embeddings.</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">94</span><span class="k">class</span> <span class="nc">LearnedPositionalEmbeddings</span><span class="p">(</span><span class="n">Module</span><span class="p">):</span></pre></div>
|
|
</div>
|
|
</div>
|
|
<div class='section' id='section-11'>
|
|
<div class='docs doc-strings'>
|
|
<div class='section-link'>
|
|
<a href='#section-11'>#</a>
|
|
</div>
|
|
<ul>
|
|
<li><code>d_model</code> is the transformer embeddings size</li>
|
|
<li><code>max_len</code> is the maximum number of patches</li>
|
|
</ul>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">103</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">max_len</span><span class="p">:</span> <span class="nb">int</span> <span class="o">=</span> <span class="mi">5_000</span><span class="p">):</span></pre></div>
|
|
</div>
|
|
</div>
|
|
<div class='section' id='section-12'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-12'>#</a>
|
|
</div>
|
|
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">108</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-13'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-13'>#</a>
|
|
</div>
|
|
<p>Positional embeddings for each location</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">110</span> <span class="bp">self</span><span class="o">.</span><span class="n">positional_encodings</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Parameter</span><span class="p">(</span><span class="n">torch</span><span class="o">.</span><span class="n">zeros</span><span class="p">(</span><span class="n">max_len</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="n">d_model</span><span class="p">),</span> <span class="n">requires_grad</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span></pre></div>
|
|
</div>
|
|
</div>
|
|
<div class='section' id='section-14'>
|
|
<div class='docs doc-strings'>
|
|
<div class='section-link'>
|
|
<a href='#section-14'>#</a>
|
|
</div>
|
|
<ul>
|
|
<li><code>x</code> is the patch embeddings of shape <code>[patches, batch_size, d_model]</code></li>
|
|
</ul>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">112</span> <span class="k">def</span> <span class="fm">__call__</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-15'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-15'>#</a>
|
|
</div>
|
|
<p>Get the positional embeddings for the given patches</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">117</span> <span class="n">pe</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">positional_encodings</span><span class="p">[</span><span class="n">x</span><span class="o">.</span><span class="n">shape</span><span class="p">[</span><span class="mi">0</span><span class="p">]]</span></pre></div>
|
|
</div>
|
|
</div>
|
|
<div class='section' id='section-16'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-16'>#</a>
|
|
</div>
|
|
<p>Add to patch embeddings and return</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">119</span> <span class="k">return</span> <span class="n">x</span> <span class="o">+</span> <span class="n">pe</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>
|
|
<p><a id="ClassificationHead"></p>
|
|
<h2>MLP Classification Head</h2>
|
|
<p></a></p>
|
|
<p>This is the two layer MLP head to classify the image based on <code>[CLS]</code> token embedding.</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">122</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-18'>
|
|
<div class='docs doc-strings'>
|
|
<div class='section-link'>
|
|
<a href='#section-18'>#</a>
|
|
</div>
|
|
<ul>
|
|
<li><code>d_model</code> is the transformer embedding size</li>
|
|
<li><code>n_hidden</code> is the size of the hidden layer</li>
|
|
<li><code>n_classes</code> is the number of classes in the classification task</li>
|
|
</ul>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">130</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_hidden</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-19'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-19'>#</a>
|
|
</div>
|
|
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">136</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-20'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-20'>#</a>
|
|
</div>
|
|
<p>First layer</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">138</span> <span class="bp">self</span><span class="o">.</span><span class="n">linear1</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_hidden</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>Activation</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">140</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">ReLU</span><span class="p">()</span></pre></div>
|
|
</div>
|
|
</div>
|
|
<div class='section' id='section-22'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-22'>#</a>
|
|
</div>
|
|
<p>Second layer</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">142</span> <span class="bp">self</span><span class="o">.</span><span class="n">linear2</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">n_hidden</span><span class="p">,</span> <span class="n">n_classes</span><span class="p">)</span></pre></div>
|
|
</div>
|
|
</div>
|
|
<div class='section' id='section-23'>
|
|
<div class='docs doc-strings'>
|
|
<div class='section-link'>
|
|
<a href='#section-23'>#</a>
|
|
</div>
|
|
<ul>
|
|
<li><code>x</code> is the transformer encoding for <code>[CLS]</code> token</li>
|
|
</ul>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">144</span> <span class="k">def</span> <span class="fm">__call__</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-24'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-24'>#</a>
|
|
</div>
|
|
<p>First layer and activation</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">149</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="bp">self</span><span class="o">.</span><span class="n">linear1</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>Second layer</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">151</span> <span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">linear2</span><span class="p">(</span><span class="n">x</span><span class="p">)</span></pre></div>
|
|
</div>
|
|
</div>
|
|
<div class='section' id='section-26'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-26'>#</a>
|
|
</div>
|
|
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">154</span> <span class="k">return</span> <span class="n">x</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>
|
|
<h2>Vision Transformer</h2>
|
|
<p>This combines the <a href="#PatchEmbeddings">patch embeddings</a>,
|
|
<a href="#LearnedPositionalEmbeddings">positional embeddings</a>,
|
|
transformer and the <a href="#ClassificationHead">classification head</a>.</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">157</span><span class="k">class</span> <span class="nc">VisionTransformer</span><span class="p">(</span><span class="n">Module</span><span class="p">):</span></pre></div>
|
|
</div>
|
|
</div>
|
|
<div class='section' id='section-28'>
|
|
<div class='docs doc-strings'>
|
|
<div class='section-link'>
|
|
<a href='#section-28'>#</a>
|
|
</div>
|
|
<ul>
|
|
<li><code>transformer_layer</code> is a copy of a single <a href="../models.html#TransformerLayer">transformer layer</a>.
|
|
We make copies of it to make the transformer with <code>n_layers</code>.</li>
|
|
<li><code>n_layers</code> is the number of [transformer layers]((../models.html#TransformerLayer).</li>
|
|
<li><code>patch_emb</code> is the <a href="#PatchEmbeddings">patch embeddings layer</a>.</li>
|
|
<li><code>pos_emb</code> is the <a href="#LearnedPositionalEmbeddings">positional embeddings layer</a>.</li>
|
|
<li><code>classification</code> is the <a href="#ClassificationHead">classification head</a>.</li>
|
|
</ul>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">165</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">transformer_layer</span><span class="p">:</span> <span class="n">TransformerLayer</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">166</span> <span class="n">patch_emb</span><span class="p">:</span> <span class="n">PatchEmbeddings</span><span class="p">,</span> <span class="n">pos_emb</span><span class="p">:</span> <span class="n">LearnedPositionalEmbeddings</span><span class="p">,</span>
|
|
<span class="lineno">167</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-29'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-29'>#</a>
|
|
</div>
|
|
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">176</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-30'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-30'>#</a>
|
|
</div>
|
|
<p>Patch embeddings</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">178</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>
|
|
<span class="lineno">179</span> <span class="bp">self</span><span class="o">.</span><span class="n">pos_emb</span> <span class="o">=</span> <span class="n">pos_emb</span></pre></div>
|
|
</div>
|
|
</div>
|
|
<div class='section' id='section-31'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-31'>#</a>
|
|
</div>
|
|
<p>Classification head</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">181</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-32'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-32'>#</a>
|
|
</div>
|
|
<p>Make copies of the transformer layer</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">183</span> <span class="bp">self</span><span class="o">.</span><span class="n">transformer_layers</span> <span class="o">=</span> <span class="n">clone_module_list</span><span class="p">(</span><span class="n">transformer_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-33'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-33'>#</a>
|
|
</div>
|
|
<p><code>[CLS]</code> token embedding</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">186</span> <span class="bp">self</span><span class="o">.</span><span class="n">cls_token_emb</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Parameter</span><span class="p">(</span><span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="n">transformer_layer</span><span class="o">.</span><span class="n">size</span><span class="p">),</span> <span class="n">requires_grad</span><span class="o">=</span><span class="kc">True</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>Final normalization layer</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">188</span> <span class="bp">self</span><span class="o">.</span><span class="n">ln</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">LayerNorm</span><span class="p">([</span><span class="n">transformer_layer</span><span class="o">.</span><span class="n">size</span><span class="p">])</span></pre></div>
|
|
</div>
|
|
</div>
|
|
<div class='section' id='section-35'>
|
|
<div class='docs doc-strings'>
|
|
<div class='section-link'>
|
|
<a href='#section-35'>#</a>
|
|
</div>
|
|
<ul>
|
|
<li><code>x</code> is the input image of shape <code>[batch_size, channels, height, width]</code></li>
|
|
</ul>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">190</span> <span class="k">def</span> <span class="fm">__call__</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-36'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-36'>#</a>
|
|
</div>
|
|
<p>Get patch embeddings. This gives a tensor of shape <code>[patches, batch_size, d_model]</code></p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">195</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-37'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-37'>#</a>
|
|
</div>
|
|
<p>Add positional embeddings</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">197</span> <span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">pos_emb</span><span class="p">(</span><span class="n">x</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>
|
|
<p>Concatenate the <code>[CLS]</code> token embeddings before feeding the transformer</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">199</span> <span class="n">cls_token_emb</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">cls_token_emb</span><span class="o">.</span><span class="n">expand</span><span class="p">(</span><span class="o">-</span><span class="mi">1</span><span class="p">,</span> <span class="n">x</span><span class="o">.</span><span class="n">shape</span><span class="p">[</span><span class="mi">1</span><span class="p">],</span> <span class="o">-</span><span class="mi">1</span><span class="p">)</span>
|
|
<span class="lineno">200</span> <span class="n">x</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">cat</span><span class="p">([</span><span class="n">cls_token_emb</span><span class="p">,</span> <span class="n">x</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>Pass through transformer layers with no attention masking</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">203</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">transformer_layers</span><span class="p">:</span>
|
|
<span class="lineno">204</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="o">=</span><span class="n">x</span><span class="p">,</span> <span class="n">mask</span><span class="o">=</span><span class="kc">None</span><span class="p">)</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>Get the transformer output of the <code>[CLS]</code> token (which is the first in the sequence).</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">207</span> <span class="n">x</span> <span class="o">=</span> <span class="n">x</span><span class="p">[</span><span class="mi">0</span><span class="p">]</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>Layer normalization</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">210</span> <span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">ln</span><span class="p">(</span><span class="n">x</span><span class="p">)</span></pre></div>
|
|
</div>
|
|
</div>
|
|
<div class='section' id='section-42'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-42'>#</a>
|
|
</div>
|
|
<p>Classification head, to get logits</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">213</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-43'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-43'>#</a>
|
|
</div>
|
|
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">216</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="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.4/MathJax.js?config=TeX-AMS_HTML">
|
|
</script>
|
|
<!-- MathJax configuration -->
|
|
<script type="text/x-mathjax-config">
|
|
MathJax.Hub.Config({
|
|
tex2jax: {
|
|
inlineMath: [ ['$','$'] ],
|
|
displayMath: [ ['$$','$$'] ],
|
|
processEscapes: true,
|
|
processEnvironments: true
|
|
},
|
|
// Center justify equations in code and markdown cells. Elsewhere
|
|
// we use CSS to left justify single line equations in code cells.
|
|
displayAlign: 'center',
|
|
"HTML-CSS": { fonts: ["TeX"] }
|
|
});
|
|
|
|
</script>
|
|
<script>
|
|
function handleImages() {
|
|
var images = document.querySelectorAll('p>img')
|
|
|
|
console.log(images);
|
|
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> |