learning/annotated_deep_learning_paper_implementations
1
0
Fork 0
mirror of https://github.com/labmlai/annotated_deep_learning_paper_implementations.git synced 2025-11-01 03:43:09 +08:00
Code Issues Packages Projects Releases Wiki Activity
Files
annotated_deep_learning_pap…/docs/transformers/rope/experiment.html
Varuna Jayasiri 05632f9f8e docs
2022-09-07 10:33:13 +05:30

452 lines
24 KiB
HTML
Raw Blame History

<!DOCTYPE html>
<html lang="en">
<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="This experiment trains a transformer model with Rotary Positional Embeddings (RoPE) on tiny Shakespeare dataset."/>
<meta name="twitter:card" content="summary"/>
<meta name="twitter:image:src" content="https://avatars1.githubusercontent.com/u/64068543?s=400&amp;v=4"/>
<meta name="twitter:title" content="Rotary Positional Embeddings (RoPE) Experiment"/>
<meta name="twitter:description" content="This experiment trains a transformer model with Rotary Positional Embeddings (RoPE) on tiny Shakespeare dataset."/>
<meta name="twitter:site" content="@labmlai"/>
<meta name="twitter:creator" content="@labmlai"/>
<meta property="og:url" content="https://nn.labml.ai/transformers/rope/experiment.html"/>
<meta property="og:title" content="Rotary Positional Embeddings (RoPE) Experiment"/>
<meta property="og:image" content="https://avatars1.githubusercontent.com/u/64068543?s=400&amp;v=4"/>
<meta property="og:site_name" content="Rotary Positional Embeddings (RoPE) Experiment"/>
<meta property="og:type" content="object"/>
<meta property="og:title" content="Rotary Positional Embeddings (RoPE) Experiment"/>
<meta property="og:description" content="This experiment trains a transformer model with Rotary Positional Embeddings (RoPE) on tiny Shakespeare dataset."/>
<title>Rotary Positional Embeddings (RoPE) Experiment</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/rope/experiment.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>
<a class="parent" href="index.html">rope</a>
</p>
<p>
<a href="https://github.com/sponsors/labmlai" target="_blank">
<img alt="Sponsor"
src="https://img.shields.io/static/v1?label=Sponsor&message=%E2%9D%A4&logo=GitHub&color=%23fe8e86"
style="max-width:100%;"/></a>
<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/rope/experiment.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>Rotary Positional Embeddings (RoPE) Experiment</h1>
<p>This is an annotated PyTorch experiment to train a transformer model with Rotary Positional Embeddings (RoPE).</p>
<p><a href="https://app.labml.ai/run/1cf508e693be11ecacc98de8b38a61fe"><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">14</span><span></span><span class="kn">from</span> <span class="nn">labml</span> <span class="kn">import</span> <span class="n">experiment</span>
<span class="lineno">15</span><span class="kn">from</span> <span class="nn">labml.configs</span> <span class="kn">import</span> <span class="n">option</span><span class="p">,</span> <span class="n">calculate</span>
<span class="lineno">16</span><span class="kn">from</span> <span class="nn">labml_nn.transformers</span> <span class="kn">import</span> <span class="n">TransformerConfigs</span>
<span class="lineno">17</span><span class="kn">from</span> <span class="nn">labml_nn.transformers.basic.autoregressive_experiment</span> <span class="kn">import</span> <span class="n">AutoregressiveTransformer</span><span class="p">,</span> <span class="n">Configs</span></pre></div>
</div>
</div>
<div class='section' id='section-1'>
<div class='docs'>
<div class='section-link'>
<a href='#section-1'>#</a>
</div>
<h3>Rotary PE attention</h3>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">21</span><span class="k">def</span> <span class="nf">_rotary_pe_mha</span><span class="p">(</span><span class="n">c</span><span class="p">:</span> <span class="n">TransformerConfigs</span><span class="p">):</span>
<span class="lineno">22</span> <span class="kn">from</span> <span class="nn">labml_nn.transformers.rope</span> <span class="kn">import</span> <span class="n">RotaryPEMultiHeadAttention</span>
<span class="lineno">23</span> <span class="k">return</span> <span class="n">RotaryPEMultiHeadAttention</span><span class="p">(</span><span class="n">c</span><span class="o">.</span><span class="n">n_heads</span><span class="p">,</span> <span class="n">c</span><span class="o">.</span><span class="n">d_model</span><span class="p">,</span> <span class="mf">1.</span><span class="p">)</span></pre></div>
</div>
</div>
<div class='section' id='section-2'>
<div class='docs'>
<div class='section-link'>
<a href='#section-2'>#</a>
</div>
<p>Configuration options </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">27</span><span class="n">calculate</span><span class="p">(</span><span class="n">TransformerConfigs</span><span class="o">.</span><span class="n">encoder_attn</span><span class="p">,</span> <span class="s1">&#39;rotary&#39;</span><span class="p">,</span> <span class="n">_rotary_pe_mha</span><span class="p">)</span>
<span class="lineno">28</span><span class="n">calculate</span><span class="p">(</span><span class="n">TransformerConfigs</span><span class="o">.</span><span class="n">decoder_attn</span><span class="p">,</span> <span class="s1">&#39;rotary&#39;</span><span class="p">,</span> <span class="n">_rotary_pe_mha</span><span class="p">)</span>
<span class="lineno">29</span><span class="n">calculate</span><span class="p">(</span><span class="n">TransformerConfigs</span><span class="o">.</span><span class="n">decoder_mem_attn</span><span class="p">,</span> <span class="s1">&#39;rotary&#39;</span><span class="p">,</span> <span class="n">_rotary_pe_mha</span><span class="p">)</span></pre></div>
</div>
</div>
<div class='section' id='section-3'>
<div class='docs doc-strings'>
<div class='section-link'>
<a href='#section-3'>#</a>
</div>
<p> Create an autoregressive model and initialize weights</p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">32</span><span class="nd">@option</span><span class="p">(</span><span class="n">Configs</span><span class="o">.</span><span class="n">model</span><span class="p">,</span> <span class="s1">&#39;rotary_pe_transformer&#39;</span><span class="p">)</span>
<span class="lineno">33</span><span class="k">def</span> <span class="nf">_model</span><span class="p">(</span><span class="n">c</span><span class="p">:</span> <span class="n">Configs</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>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">37</span> <span class="n">m</span> <span class="o">=</span> <span class="n">AutoregressiveTransformer</span><span class="p">(</span><span class="n">c</span><span class="o">.</span><span class="n">transformer</span><span class="o">.</span><span class="n">encoder</span><span class="p">,</span>
<span class="lineno">38</span> <span class="n">c</span><span class="o">.</span><span class="n">transformer</span><span class="o">.</span><span class="n">src_embed</span><span class="p">,</span>
<span class="lineno">39</span> <span class="n">c</span><span class="o">.</span><span class="n">transformer</span><span class="o">.</span><span class="n">generator</span><span class="p">)</span><span class="o">.</span><span class="n">to</span><span class="p">(</span><span class="n">c</span><span class="o">.</span><span class="n">device</span><span class="p">)</span>
<span class="lineno">40</span>
<span class="lineno">41</span> <span class="k">return</span> <span class="n">m</span></pre></div>
</div>
</div>
<div class='section' id='section-5'>
<div class='docs'>
<div class='section-link'>
<a href='#section-5'>#</a>
</div>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">44</span><span class="k">def</span> <span class="nf">main</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>Create experiment </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">46</span> <span class="n">experiment</span><span class="o">.</span><span class="n">create</span><span class="p">(</span><span class="n">name</span><span class="o">=</span><span class="s2">&quot;rotary_pe_transformer&quot;</span><span class="p">,</span> <span class="n">writers</span><span class="o">=</span><span class="p">{</span><span class="s1">&#39;screen&#39;</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>Create configs </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">48</span> <span class="n">conf</span> <span class="o">=</span> <span class="n">Configs</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>Override configurations </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">50</span> <span class="n">experiment</span><span class="o">.</span><span class="n">configs</span><span class="p">(</span><span class="n">conf</span><span class="p">,</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>No fixed positional embeddings </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">52</span> <span class="s1">&#39;transformer.src_embed&#39;</span><span class="p">:</span> <span class="s1">&#39;no_pos&#39;</span><span class="p">,</span>
<span class="lineno">53</span> <span class="s1">&#39;transformer.tgt_embed&#39;</span><span class="p">:</span> <span class="s1">&#39;no_pos&#39;</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>
<p>Encoder with RoPE </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">56</span> <span class="s1">&#39;transformer.encoder_attn&#39;</span><span class="p">:</span> <span class="s1">&#39;rotary&#39;</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> </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">59</span> <span class="s1">&#39;model&#39;</span><span class="p">:</span> <span class="s1">&#39;rotary_pe_transformer&#39;</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>
<p>Use character level tokenizer </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">62</span> <span class="s1">&#39;tokenizer&#39;</span><span class="p">:</span> <span class="s1">&#39;character&#39;</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>Prompt separator is blank </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">64</span> <span class="s1">&#39;prompt_separator&#39;</span><span class="p">:</span> <span class="s1">&#39;&#39;</span><span class="p">,</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>Starting prompt for sampling </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">66</span> <span class="s1">&#39;prompt&#39;</span><span class="p">:</span> <span class="s1">&#39;It is &#39;</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>Use Tiny Shakespeare dataset </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">68</span> <span class="s1">&#39;text&#39;</span><span class="p">:</span> <span class="s1">&#39;tiny_shakespeare&#39;</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>Use a context size of <span ><span class="katex"><span aria-hidden="true" class="katex-html"><span class="base"><span class="strut" style="height:0.64444em;vertical-align:0em;"></span><span class="mord">256</span></span></span></span></span> </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">71</span> <span class="s1">&#39;seq_len&#39;</span><span class="p">:</span> <span class="mi">512</span><span class="p">,</span></pre></div>
</div>
</div>
<div class='section' id='section-17'>
<div class='docs'>
<div class='section-link'>
<a href='#section-17'>#</a>
</div>
<p>Train for 32 epochs </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">73</span> <span class="s1">&#39;epochs&#39;</span><span class="p">:</span> <span class="mi">32</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>
<p>Batch size <span ><span class="katex"><span aria-hidden="true" class="katex-html"><span class="base"><span class="strut" style="height:0.64444em;vertical-align:0em;"></span><span class="mord">4</span></span></span></span></span> </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">75</span> <span class="s1">&#39;batch_size&#39;</span><span class="p">:</span> <span class="mi">4</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>Switch between training and validation for <span ><span class="katex"><span aria-hidden="true" class="katex-html"><span class="base"><span class="strut" style="height:0.64444em;vertical-align:0em;"></span><span class="mord">10</span></span></span></span></span> times per epoch </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">78</span> <span class="s1">&#39;inner_iterations&#39;</span><span class="p">:</span> <span class="mi">10</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>Model size </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">81</span> <span class="s1">&#39;d_model&#39;</span><span class="p">:</span> <span class="mi">128</span><span class="p">,</span>
<span class="lineno">82</span> <span class="s1">&#39;transformer.ffn.d_ff&#39;</span><span class="p">:</span> <span class="mi">512</span><span class="p">,</span>
<span class="lineno">83</span> <span class="s1">&#39;transformer.n_heads&#39;</span><span class="p">:</span> <span class="mi">16</span><span class="p">,</span>
<span class="lineno">84</span> <span class="s1">&#39;transformer.dropout&#39;</span><span class="p">:</span> <span class="mf">0.0</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>Use <a href="../../optimizers/noam.html">Noam optimizer</a> </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">87</span> <span class="s1">&#39;optimizer.optimizer&#39;</span><span class="p">:</span> <span class="s1">&#39;Noam&#39;</span><span class="p">,</span>
<span class="lineno">88</span> <span class="s1">&#39;optimizer.learning_rate&#39;</span><span class="p">:</span> <span class="mf">1.</span><span class="p">,</span>
<span class="lineno">89</span>
<span class="lineno">90</span> <span class="s1">&#39;dataloader_shuffle_with_replacement&#39;</span><span class="p">:</span> <span class="kc">True</span>
<span class="lineno">91</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>Set models for saving and loading </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">94</span> <span class="n">experiment</span><span class="o">.</span><span class="n">add_pytorch_models</span><span class="p">({</span><span class="s1">&#39;model&#39;</span><span class="p">:</span> <span class="n">conf</span><span class="o">.</span><span class="n">model</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>Start the experiment </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">97</span> <span class="k">with</span> <span class="n">experiment</span><span class="o">.</span><span class="n">start</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>Run training </p>
</div>
<div class='code'>
<div class="highlight"><pre><span class="lineno">99</span> <span class="n">conf</span><span class="o">.</span><span class="n">run</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">103</span><span class="k">if</span> <span class="vm">__name__</span> <span class="o">==</span> <span class="s1">&#39;__main__&#39;</span><span class="p">:</span>
<span class="lineno">104</span> <span class="n">main</span><span class="p">()</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>
Reference in New Issue View Git Blame Copy Permalink