mirror of
https://github.com/labmlai/annotated_deep_learning_paper_implementations.git
synced 2025-10-31 02:39:16 +08:00
171 lines
9.8 KiB
HTML
171 lines
9.8 KiB
HTML
<!DOCTYPE html>
|
||
<html lang="zh">
|
||
<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="这是一个包含 Transformers 及相关技术的 PyTorch 实现和教程的合集。"/>
|
||
|
||
<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="Transformers"/>
|
||
<meta name="twitter:description" content="这是一个包含 Transformers 及相关技术的 PyTorch 实现和教程的合集。"/>
|
||
<meta name="twitter:site" content="@labmlai"/>
|
||
<meta name="twitter:creator" content="@labmlai"/>
|
||
|
||
<meta property="og:url" content="https://nn.labml.ai/transformers/index.html"/>
|
||
<meta property="og:title" content="Transformers"/>
|
||
<meta property="og:image" content="https://avatars1.githubusercontent.com/u/64068543?s=400&v=4"/>
|
||
<meta property="og:site_name" content="Transformers"/>
|
||
<meta property="og:type" content="object"/>
|
||
<meta property="og:title" content="Transformers"/>
|
||
<meta property="og:description" content="这是一个包含 Transformers 及相关技术的 PyTorch 实现和教程的合集。"/>
|
||
|
||
<title>Transformers</title>
|
||
<link rel="shortcut icon" href="/icon.png"/>
|
||
<link rel="stylesheet" href="../pylit.css?v=1">
|
||
<link rel="canonical" href="https://nn.labml.ai/transformers/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">transformers</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/transformers/__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>Transformers</h1>
|
||
</a><p>本节内容包含对论文<a href="https://arxiv.org/abs/1706.03762">《 Attention is All You Need 》</a>中原始 Transformer 的解释与<a href="https://pytorch.org/">PyTorch</a> 实现，以及对其衍生和增强版本的解释与实现。</p>
|
||
<ul><li><a href="mha.html">多头注意力</a></li>
|
||
<li><a href="models.html">Transformer 编码器和解码器模型</a></li>
|
||
<li><a href="feed_forward.html">位置前馈网络 (FFN)</a></li>
|
||
<li><a href="positional_encoding.html">固定位置编码</a></li></ul>
|
||
<h2><a href="xl/index.html">Transformer XL</a></h2>
|
||
<p>这是使用<a href="xl/relative_mha.html">相对多头注意力</a>的 Transformer XL 模型的实现。</p>
|
||
<h2><a href="rope/index.html">旋转式位置编码</a></h2>
|
||
<p>这是旋转式位置编码（ ROPE ）的实现。</p>
|
||
<h2><a href="alibi/index.html">线性偏差注意力</a></h2>
|
||
<p>这是线性偏差注意力（ ALIBI ）的实现。</p>
|
||
<h2><a href="retro/index.html">RETRO</a></h2>
|
||
<p>这是对检索增强 Transformer （ RETRO ）的实现。</p>
|
||
<h2><a href="compressive/index.html">压缩 Transformer</a></h2>
|
||
<p>这是一个压缩transformer的实现，它在<a href="xl/index.html">Transformer XL</a> 的基础上，通过压缩最早期的记忆来延长注意力跨度。</p>
|
||
<h2><a href="gpt/index.html">GPT 架构</a></h2>
|
||
<p>这是 GPT-2 结构的实现。</p>
|
||
<h2><a href="glu_variants/simple.html">GLU 变体</a></h2>
|
||
<p>这是论文 <a href="https://arxiv.org/abs/2002.05202">《 GLU Variants Improve Transformer 》</a>的实现。</p>
|
||
<h2><a href="knn/index.html">kNN-LM</a></h2>
|
||
<p>这是论文<a href="https://arxiv.org/abs/1911.00172">《 Generalization through Memorization: Nearest Neighbor Language Models 》</a>的实现。</p>
|
||
<h2><a href="feedback/index.html">自反馈 Transformer</a></h2>
|
||
<p>这是论文<a href="https://arxiv.org/abs/2002.09402">《 Accessing Higher-level Representations in Sequential Transformers with Feedback Memory 》</a>的实现。</p>
|
||
<h2><a href="switch/index.html">Switch Transformer</a></h2>
|
||
<p>这是论文<a href="https://arxiv.org/abs/2101.03961">《 Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity 》</a>的一个简化实现。我们的实现仅包含几百万个参数，并且只在单 GPU 上进行训练，不涉及并行分布式训练，但我们仍然实现了论文中描述的 Switch 概念。</p>
|
||
<h2><a href="fast_weights/index.html">快速权重 Transformer</a></h2>
|
||
<p>这是论文 <a href="https://arxiv.org/abs/2102.11174">《 Linear Transformers Are Secretly Fast Weight Memory Systems in PyTorch 》</a>的实现。</p>
|
||
<h2><a href="fnet/index.html">Fnet：使用傅里叶变换混合 token </a></h2>
|
||
<p>这是论文<a href="https://arxiv.org/abs/2105.03824">《 FNet: Mixing Tokens with Fourier Transforms 》</a>的实现。</p>
|
||
<h2><a href="aft/index.html">无注意力 Transformer</a></h2>
|
||
<p>这是论文<a href="https://arxiv.org/abs/2105.14103">《 An Attention Free Transformer 》</a>的实现。</p>
|
||
<h2><a href="mlm/index.html">掩码语言模型</a></h2>
|
||
<p>这是论文<a href="https://arxiv.org/abs/1810.04805">《 BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding 》</a>中用于预训练的掩码语言模型的实现</p>
|
||
<h2><a href="mlp_mixer/index.html">MLP-Mixer：一种用于视觉的全 MLP 架构</a></h2>
|
||
<p>这是论文 <a href="https://arxiv.org/abs/2105.01601">《 MLP-Mixer: An all-MLP Architecture for Vision 》</a>的实现。</p>
|
||
<h2><a href="gmlp/index.html">门控多层感知器 (gMLP)</a></h2>
|
||
<p>这是论文<a href="https://arxiv.org/abs/2105.08050">《 Pay Attention to MLPs 》</a>的实现。</p>
|
||
<h2><a href="vit/index.html">视觉 Transformer (ViT)</a></h2>
|
||
<p>这是论文<a href="https://arxiv.org/abs/2010.11929">《 An Image Is Worth 16x16 Words: Transformers For Image Recognition At Scale 》</a>的实现。</p>
|
||
<h2><a href="primer_ez/index.html">Primer</a></h2>
|
||
<p>这是论文<a href="https://arxiv.org/abs/2109.08668">《 Primer: Searching for Efficient Transformers for Language Modeling 》</a>的实现。</p>
|
||
<h2><a href="hour_glass/index.html">沙漏网络</a></h2>
|
||
<p>这是论文<a href="https://arxiv.org/abs/2110.13711">《 Hierarchical Transformers Are More Efficient Language Models 》</a>的实现</p>
|
||
|
||
</div>
|
||
<div class='code'>
|
||
<div class="highlight"><pre><span class="lineno">112</span><span></span><span class="kn">from</span> <span class="nn">.configs</span> <span class="kn">import</span> <span class="n">TransformerConfigs</span>
|
||
<span class="lineno">113</span><span class="kn">from</span> <span class="nn">.models</span> <span class="kn">import</span> <span class="n">TransformerLayer</span><span class="p">,</span> <span class="n">Encoder</span><span class="p">,</span> <span class="n">Decoder</span><span class="p">,</span> <span class="n">Generator</span><span class="p">,</span> <span class="n">EncoderDecoder</span>
|
||
<span class="lineno">114</span><span class="kn">from</span> <span class="nn">.mha</span> <span class="kn">import</span> <span class="n">MultiHeadAttention</span>
|
||
<span class="lineno">115</span><span class="kn">from</span> <span class="nn">labml_nn.transformers.xl.relative_mha</span> <span class="kn">import</span> <span class="n">RelativeMultiHeadAttention</span></pre></div>
|
||
</div>
|
||
</div>
|
||
<div class='footer'>
|
||
<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> |