mirror of
https://github.com/labmlai/annotated_deep_learning_paper_implementations.git
synced 2025-11-03 22:09:18 +08:00
291 lines
24 KiB
HTML
291 lines
24 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=""/>
|
|
|
|
<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="ray_tune.py"/>
|
|
<meta name="twitter:description" content=""/>
|
|
<meta name="twitter:site" content="@labmlai"/>
|
|
<meta name="twitter:creator" content="@labmlai"/>
|
|
|
|
<meta property="og:url" content="https://nn.labml.ai/cnn/ray_tune.html"/>
|
|
<meta property="og:title" content="ray_tune.py"/>
|
|
<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="ray_tune.py"/>
|
|
<meta property="og:description" content=""/>
|
|
|
|
<title>ray_tune.py</title>
|
|
<link rel="shortcut icon" href="/icon.png"/>
|
|
<link rel="stylesheet" href="../pylit.css">
|
|
<link rel="canonical" href="https://nn.labml.ai/cnn/ray_tune.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">cnn</a>
|
|
</p>
|
|
<p>
|
|
|
|
<a href="https://github.com/lab-ml/labml_nn/tree/master/labml_nn/cnn/ray_tune.py">
|
|
<img alt="Github"
|
|
src="https://img.shields.io/github/stars/lab-ml/nn?style=social"
|
|
style="max-width:100%;"/></a>
|
|
<a href="https://join.slack.com/t/labforml/shared_invite/zt-egj9zvq9-Dl3hhZqobexgT7aVKnD14g/"
|
|
rel="nofollow">
|
|
<img alt="Join Slact"
|
|
src="https://img.shields.io/badge/slack-chat-green.svg?logo=slack"
|
|
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'>
|
|
<div class='section-link'>
|
|
<a href='#section-0'>#</a>
|
|
</div>
|
|
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">3</span><span></span><span class="kn">import</span> <span class="nn">numpy</span> <span class="k">as</span> <span class="nn">np</span>
|
|
<span class="lineno">4</span><span class="kn">import</span> <span class="nn">os</span>
|
|
<span class="lineno">5</span><span class="kn">import</span> <span class="nn">torch</span>
|
|
<span class="lineno">6</span><span class="kn">from</span> <span class="nn">ray</span> <span class="kn">import</span> <span class="n">tune</span>
|
|
<span class="lineno">7</span><span class="kn">from</span> <span class="nn">ray.tune.schedulers</span> <span class="kn">import</span> <span class="n">ASHAScheduler</span><span class="p">,</span> <span class="n">PopulationBasedTraining</span>
|
|
<span class="lineno">8</span><span class="kn">from</span> <span class="nn">utils.train</span> <span class="kn">import</span> <span class="n">Trainer</span>
|
|
<span class="lineno">9</span><span class="kn">from</span> <span class="nn">models.cnn</span> <span class="kn">import</span> <span class="n">GetCNN</span></pre></div>
|
|
</div>
|
|
</div>
|
|
<div class='section' id='section-1'>
|
|
<div class='docs'>
|
|
<div class='section-link'>
|
|
<a href='#section-1'>#</a>
|
|
</div>
|
|
<p>Check if GPU is available</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">12</span><span class="n">device</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">device</span><span class="p">(</span><span class="s2">"cuda:0"</span> <span class="k">if</span> <span class="n">torch</span><span class="o">.</span><span class="n">cuda</span><span class="o">.</span><span class="n">is_available</span><span class="p">()</span> <span class="k">else</span> <span class="s2">"cpu"</span><span class="p">)</span>
|
|
<span class="lineno">13</span><span class="nb">print</span><span class="p">(</span><span class="s2">"Device: "</span> <span class="o">+</span> <span class="nb">str</span><span class="p">(</span><span class="n">device</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>
|
|
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">16</span><span class="n">num_samples</span><span class="o">=</span> <span class="mi">40</span> <span class="c1"># for multiple trials</span>
|
|
<span class="lineno">17</span><span class="n">max_num_epochs</span><span class="o">=</span> <span class="mi">25</span>
|
|
<span class="lineno">18</span><span class="n">gpus_per_trial</span><span class="o">=</span> <span class="mi">1</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>Cifar 10 Datasets location</p>
|
|
<p>Code has been referenced from the official ray tune documentation
|
|
ASHA
|
|
https://docs.ray.io/en/master/tune/api_docs/schedulers.html#tune-scheduler-hyperband</p>
|
|
<p>PBT
|
|
https://docs.ray.io/en/latest/tune/api_docs/schedulers.html#tune-scheduler-pbt</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">21</span><span class="n">data_dir</span> <span class="o">=</span> <span class="s1">'./data/Cifar10'</span></pre></div>
|
|
</div>
|
|
</div>
|
|
<div class='section' id='section-4'>
|
|
<div class='docs doc-strings'>
|
|
<div class='section-link'>
|
|
<a href='#section-4'>#</a>
|
|
</div>
|
|
<p>config - returns a dict of hyperparameters</p>
|
|
<p>Selecting different hyperparameters for tuning
|
|
l1 : Number of units in first fully connected layer
|
|
l2 : Number of units in second fully connected layer
|
|
lr : Learning rate
|
|
decay : Decay rate for regularization
|
|
batch_size : Batch size of test and train data</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre></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">41</span><span class="n">config</span> <span class="o">=</span> <span class="p">{</span>
|
|
<span class="lineno">42</span> <span class="s2">"l1"</span><span class="p">:</span> <span class="n">tune</span><span class="o">.</span><span class="n">sample_from</span><span class="p">(</span><span class="k">lambda</span> <span class="n">_</span><span class="p">:</span> <span class="mi">2</span> <span class="o">**</span> <span class="n">np</span><span class="o">.</span><span class="n">random</span><span class="o">.</span><span class="n">randint</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">9</span><span class="p">)),</span> <span class="c1"># eg. 4, 8, 16 .. 512</span>
|
|
<span class="lineno">43</span> <span class="s2">"l2"</span><span class="p">:</span> <span class="n">tune</span><span class="o">.</span><span class="n">sample_from</span><span class="p">(</span><span class="k">lambda</span> <span class="n">_</span><span class="p">:</span> <span class="mi">2</span> <span class="o">**</span> <span class="n">np</span><span class="o">.</span><span class="n">random</span><span class="o">.</span><span class="n">randint</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">9</span><span class="p">)),</span> <span class="c1"># eg. 4, 8, 16 .. 512</span>
|
|
<span class="lineno">44</span> <span class="s2">"lr"</span><span class="p">:</span> <span class="n">tune</span><span class="o">.</span><span class="n">loguniform</span><span class="p">(</span><span class="mf">1e-4</span><span class="p">,</span> <span class="mf">1e-1</span><span class="p">),</span> <span class="c1"># Sampling from log uniform distribution</span>
|
|
<span class="lineno">45</span> <span class="s2">"decay"</span><span class="p">:</span> <span class="n">tune</span><span class="o">.</span><span class="n">sample_from</span><span class="p">(</span><span class="k">lambda</span> <span class="n">_</span><span class="p">:</span> <span class="mi">10</span> <span class="o">**</span> <span class="n">np</span><span class="o">.</span><span class="n">random</span><span class="o">.</span><span class="n">randint</span><span class="p">(</span><span class="o">-</span><span class="mi">7</span><span class="p">,</span> <span class="o">-</span><span class="mi">3</span><span class="p">)),</span> <span class="c1"># eg. 1e-7, 1e-6, .. 1e-3</span>
|
|
<span class="lineno">46</span> <span class="s2">"batch_size"</span><span class="p">:</span> <span class="n">tune</span><span class="o">.</span><span class="n">choice</span><span class="p">([</span><span class="mi">32</span><span class="p">,</span> <span class="mi">64</span><span class="p">,</span> <span class="mi">128</span><span class="p">,</span> <span class="mi">256</span><span class="p">])</span>
|
|
<span class="lineno">47</span><span class="p">}</span></pre></div>
|
|
</div>
|
|
</div>
|
|
<div class='section' id='section-6'>
|
|
<div class='docs doc-strings'>
|
|
<div class='section-link'>
|
|
<a href='#section-6'>#</a>
|
|
</div>
|
|
<p>calling trainer
|
|
ASHA (Asynchronous Successive Halving Algorithm) scheduler
|
|
max_t : Maximum number of units per trail (can be time or epochs)
|
|
grace_period : Stop trials after specific number of unit if model is not performing well (can be time or epochs)
|
|
reduction_factor : Set halving rate</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">50</span><span class="n">trainer</span> <span class="o">=</span> <span class="n">Trainer</span><span class="p">(</span><span class="n">device</span><span class="o">=</span><span class="n">device</span><span class="p">)</span></pre></div>
|
|
</div>
|
|
</div>
|
|
<div class='section' id='section-7'>
|
|
<div class='docs doc-strings'>
|
|
<div class='section-link'>
|
|
<a href='#section-7'>#</a>
|
|
</div>
|
|
<p>Population based training scheduler
|
|
time_attr : Can be time or epochs
|
|
metric : Objective of training (loss or accuracy)
|
|
perturbation_interval : Perturbation occur after specified unit (can be time or epochs)
|
|
hyperparam_mutations : Hyperparameters to mutate</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">57</span><span class="n">scheduler</span> <span class="o">=</span> <span class="n">ASHAScheduler</span><span class="p">(</span>
|
|
<span class="lineno">58</span> <span class="n">max_t</span><span class="o">=</span><span class="n">max_num_epochs</span><span class="p">,</span>
|
|
<span class="lineno">59</span> <span class="n">grace_period</span><span class="o">=</span><span class="mi">4</span><span class="p">,</span>
|
|
<span class="lineno">60</span> <span class="n">reduction_factor</span><span class="o">=</span><span class="mi">4</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>
|
|
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">70</span><span class="n">scheduler</span> <span class="o">=</span> <span class="n">PopulationBasedTraining</span><span class="p">(</span>
|
|
<span class="lineno">71</span> <span class="n">time_attr</span><span class="o">=</span> <span class="s2">"training_iteration"</span><span class="p">,</span> <span class="c1"># epochs</span>
|
|
<span class="lineno">72</span> <span class="n">metric</span><span class="o">=</span><span class="s1">'loss'</span><span class="p">,</span> <span class="c1"># loss is objective function</span>
|
|
<span class="lineno">73</span> <span class="n">mode</span><span class="o">=</span><span class="s1">'min'</span><span class="p">,</span> <span class="c1"># minimizing loss is objective of training</span>
|
|
<span class="lineno">74</span> <span class="n">perturbation_interval</span><span class="o">=</span><span class="mf">5.0</span><span class="p">,</span> <span class="c1"># after 5 epochs perturbate</span>
|
|
<span class="lineno">75</span> <span class="n">hyperparam_mutations</span><span class="o">=</span><span class="p">{</span>
|
|
<span class="lineno">76</span> <span class="s2">"lr"</span><span class="p">:</span> <span class="p">[</span><span class="mf">1e-3</span><span class="p">,</span> <span class="mf">5e-4</span><span class="p">,</span> <span class="mf">1e-4</span><span class="p">,</span> <span class="mf">5e-4</span><span class="p">,</span> <span class="mf">1e-5</span><span class="p">],</span> <span class="c1"># choose from given learning rates</span>
|
|
<span class="lineno">77</span> <span class="s2">"batch_size"</span><span class="p">:</span> <span class="p">[</span><span class="mi">64</span><span class="p">,</span> <span class="mi">128</span><span class="p">,</span> <span class="mi">256</span><span class="p">],</span> <span class="c1"># choose from given batch sizes</span>
|
|
<span class="lineno">78</span> <span class="s2">"decay"</span><span class="p">:</span> <span class="n">tune</span><span class="o">.</span><span class="n">uniform</span><span class="p">(</span><span class="mi">10</span><span class="o">**-</span><span class="mi">8</span><span class="p">,</span> <span class="mi">10</span><span class="o">**-</span><span class="mi">4</span><span class="p">)</span> <span class="c1"># sample from uniform distribution</span>
|
|
<span class="lineno">79</span> <span class="p">}</span>
|
|
<span class="lineno">80</span> <span class="p">)</span>
|
|
<span class="lineno">81</span>
|
|
<span class="lineno">82</span><span class="n">result</span> <span class="o">=</span> <span class="n">tune</span><span class="o">.</span><span class="n">run</span><span class="p">(</span>
|
|
<span class="lineno">83</span> <span class="n">tune</span><span class="o">.</span><span class="n">with_parameters</span><span class="p">(</span><span class="n">trainer</span><span class="o">.</span><span class="n">Train_ray</span><span class="p">,</span> <span class="n">data_dir</span><span class="o">=</span><span class="n">data_dir</span><span class="p">),</span>
|
|
<span class="lineno">84</span> <span class="n">name</span><span class="o">=</span><span class="s2">"ray_test_basic-CNN"</span><span class="p">,</span> <span class="c1"># name for identifying models (checkpoints)</span>
|
|
<span class="lineno">85</span> <span class="n">scheduler</span><span class="o">=</span><span class="n">scheduler</span><span class="p">,</span> <span class="c1"># select scheduler PBT or ASHA</span>
|
|
<span class="lineno">86</span> <span class="n">resources_per_trial</span><span class="o">=</span><span class="p">{</span><span class="s2">"cpu"</span><span class="p">:</span> <span class="mi">8</span><span class="p">,</span> <span class="s2">"gpu"</span><span class="p">:</span> <span class="n">gpus_per_trial</span><span class="p">},</span> <span class="c1"># select number of CPUs or GPUs</span>
|
|
<span class="lineno">87</span> <span class="n">config</span><span class="o">=</span><span class="n">config</span><span class="p">,</span> <span class="c1"># input config dict consisting of different hyperparameters</span>
|
|
<span class="lineno">88</span> <span class="n">stop</span><span class="o">=</span><span class="p">{</span>
|
|
<span class="lineno">89</span> <span class="s2">"training_iteration"</span><span class="p">:</span> <span class="n">max_num_epochs</span><span class="p">,</span> <span class="c1"># stopping criterea</span>
|
|
<span class="lineno">90</span> <span class="p">},</span>
|
|
<span class="lineno">91</span> <span class="n">metric</span><span class="o">=</span><span class="s2">"loss"</span><span class="p">,</span> <span class="c1"># uncomment for ASHA scheduler</span>
|
|
<span class="lineno">92</span> <span class="n">mode</span><span class="o">=</span><span class="s2">"min"</span><span class="p">,</span> <span class="c1"># uncomment for ASHA scheduler</span>
|
|
<span class="lineno">93</span> <span class="n">num_samples</span><span class="o">=</span><span class="n">num_samples</span><span class="p">,</span>
|
|
<span class="lineno">94</span> <span class="n">verbose</span><span class="o">=</span><span class="kc">True</span><span class="p">,</span> <span class="c1"># keep to true to check how training progresses</span>
|
|
<span class="lineno">95</span> <span class="n">fail_fast</span><span class="o">=</span><span class="kc">True</span><span class="p">,</span> <span class="c1"># fail on first error</span>
|
|
<span class="lineno">96</span> <span class="n">keep_checkpoints_num</span><span class="o">=</span><span class="mi">5</span><span class="p">,</span> <span class="c1"># number of checkpoints to be saved per num_samples</span>
|
|
<span class="lineno">97</span>
|
|
<span class="lineno">98</span><span class="p">)</span>
|
|
<span class="lineno">99</span>
|
|
<span class="lineno">100</span><span class="n">best_trial</span> <span class="o">=</span> <span class="n">result</span><span class="o">.</span><span class="n">get_best_trial</span><span class="p">(</span><span class="s2">"loss"</span><span class="p">,</span> <span class="s2">"min"</span><span class="p">,</span> <span class="s2">"last"</span><span class="p">)</span>
|
|
<span class="lineno">101</span><span class="nb">print</span><span class="p">(</span><span class="s2">"Best configuration: </span><span class="si">{}</span><span class="s2">"</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">best_trial</span><span class="o">.</span><span class="n">config</span><span class="p">))</span>
|
|
<span class="lineno">102</span><span class="nb">print</span><span class="p">(</span><span class="s2">"Best validation loss: </span><span class="si">{}</span><span class="s2">"</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">best_trial</span><span class="o">.</span><span class="n">last_result</span><span class="p">[</span><span class="s2">"loss"</span><span class="p">]))</span>
|
|
<span class="lineno">103</span><span class="nb">print</span><span class="p">(</span><span class="s2">"Best validation accuracy: </span><span class="si">{}</span><span class="s2">"</span><span class="o">.</span><span class="n">format</span><span class="p">(</span>
|
|
<span class="lineno">104</span> <span class="n">best_trial</span><span class="o">.</span><span class="n">last_result</span><span class="p">[</span><span class="s2">"accuracy"</span><span class="p">]))</span>
|
|
<span class="lineno">105</span>
|
|
<span class="lineno">106</span>
|
|
<span class="lineno">107</span><span class="n">best_trained_model</span> <span class="o">=</span> <span class="n">GetCNN</span><span class="p">(</span><span class="n">best_trial</span><span class="o">.</span><span class="n">config</span><span class="p">[</span><span class="s2">"l1"</span><span class="p">],</span> <span class="n">best_trial</span><span class="o">.</span><span class="n">config</span><span class="p">[</span><span class="s2">"l2"</span><span class="p">])</span>
|
|
<span class="lineno">108</span><span class="n">best_trained_model</span><span class="o">.</span><span class="n">to</span><span class="p">(</span><span class="n">device</span><span class="p">)</span>
|
|
<span class="lineno">109</span><span class="n">checkpoint_path</span> <span class="o">=</span> <span class="n">os</span><span class="o">.</span><span class="n">path</span><span class="o">.</span><span class="n">join</span><span class="p">(</span><span class="n">best_trial</span><span class="o">.</span><span class="n">checkpoint</span><span class="o">.</span><span class="n">value</span><span class="p">,</span> <span class="s2">"checkpoint"</span><span class="p">)</span>
|
|
<span class="lineno">110</span><span class="n">model_state</span><span class="p">,</span> <span class="n">optimizer_state</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">load</span><span class="p">(</span><span class="n">checkpoint_path</span><span class="p">)</span>
|
|
<span class="lineno">111</span><span class="n">best_trained_model</span><span class="o">.</span><span class="n">load_state_dict</span><span class="p">(</span><span class="n">model_state</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>Check accuracy of best model</p>
|
|
</div>
|
|
<div class='code'>
|
|
<div class="highlight"><pre><span class="lineno">114</span><span class="n">test_acc</span> <span class="o">=</span> <span class="n">trainer</span><span class="o">.</span><span class="n">Test</span><span class="p">(</span><span class="n">best_trained_model</span><span class="p">,</span> <span class="n">save</span><span class="o">=</span><span class="n">data_dir</span><span class="p">)</span>
|
|
<span class="lineno">115</span><span class="nb">print</span><span class="p">(</span><span class="s2">"Best Test accuracy: </span><span class="si">{}</span><span class="s2">"</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">test_acc</span><span class="p">))</span></pre></div>
|
|
</div>
|
|
</div>
|
|
</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>
|
|
</body>
|
|
</html> |