deepG / Git / [1c0e03] /docs/articles/using

Models:
MarcoTheBlack/
deepG
Downloads: 2
[1c0e03]: / docs / articles / using_tb.html
History
Download this file
279 lines (245 with data), 26.2 kB

<!DOCTYPE html>
<!-- Generated by pkgdown: do not edit by hand --><html lang="en">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<meta charset="utf-8">
<meta http-equiv="X-UA-Compatible" content="IE=edge">
<meta name="viewport" content="width=device-width, initial-scale=1, shrink-to-fit=no">
<meta name="description" content="deepG">
<title>Using Tensorboard • deepG</title>
<!-- favicons --><link rel="icon" type="image/png" sizes="16x16" href="../favicon-16x16.png">
<link rel="icon" type="image/png" sizes="32x32" href="../favicon-32x32.png">
<link rel="apple-touch-icon" type="image/png" sizes="180x180" href="../apple-touch-icon.png">
<link rel="apple-touch-icon" type="image/png" sizes="120x120" href="../apple-touch-icon-120x120.png">
<link rel="apple-touch-icon" type="image/png" sizes="76x76" href="../apple-touch-icon-76x76.png">
<link rel="apple-touch-icon" type="image/png" sizes="60x60" href="../apple-touch-icon-60x60.png">
<script src="../deps/jquery-3.6.0/jquery-3.6.0.min.js"></script><meta name="viewport" content="width=device-width, initial-scale=1, shrink-to-fit=no">
<link href="../deps/bootstrap-5.3.1/bootstrap.min.css" rel="stylesheet">
<script src="../deps/bootstrap-5.3.1/bootstrap.bundle.min.js"></script><!-- Font Awesome icons --><link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/5.12.1/css/all.min.css" integrity="sha256-mmgLkCYLUQbXn0B1SRqzHar6dCnv9oZFPEC1g1cwlkk=" crossorigin="anonymous">
<link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/5.12.1/css/v4-shims.min.css" integrity="sha256-wZjR52fzng1pJHwx4aV2AO3yyTOXrcDW7jBpJtTwVxw=" crossorigin="anonymous">
<!-- bootstrap-toc --><script src="https://cdn.jsdelivr.net/gh/afeld/bootstrap-toc@v1.0.1/dist/bootstrap-toc.min.js" integrity="sha256-4veVQbu7//Lk5TSmc7YV48MxtMy98e26cf5MrgZYnwo=" crossorigin="anonymous"></script><!-- headroom.js --><script src="https://cdnjs.cloudflare.com/ajax/libs/headroom/0.11.0/headroom.min.js" integrity="sha256-AsUX4SJE1+yuDu5+mAVzJbuYNPHj/WroHuZ8Ir/CkE0=" crossorigin="anonymous"></script><script src="https://cdnjs.cloudflare.com/ajax/libs/headroom/0.11.0/jQuery.headroom.min.js" integrity="sha256-ZX/yNShbjqsohH1k95liqY9Gd8uOiE1S4vZc+9KQ1K4=" crossorigin="anonymous"></script><!-- clipboard.js --><script src="https://cdnjs.cloudflare.com/ajax/libs/clipboard.js/2.0.11/clipboard.min.js" integrity="sha512-7O5pXpc0oCRrxk8RUfDYFgn0nO1t+jLuIOQdOMRp4APB7uZ4vSjspzp5y6YDtDs4VzUSTbWzBFZ/LKJhnyFOKw==" crossorigin="anonymous" referrerpolicy="no-referrer"></script><!-- search --><script src="https://cdnjs.cloudflare.com/ajax/libs/fuse.js/6.4.6/fuse.js" integrity="sha512-zv6Ywkjyktsohkbp9bb45V6tEMoWhzFzXis+LrMehmJZZSys19Yxf1dopHx7WzIKxr5tK2dVcYmaCk2uqdjF4A==" crossorigin="anonymous"></script><script src="https://cdnjs.cloudflare.com/ajax/libs/autocomplete.js/0.38.0/autocomplete.jquery.min.js" integrity="sha512-GU9ayf+66Xx2TmpxqJpliWbT5PiGYxpaG8rfnBEk1LL8l1KGkRShhngwdXK1UgqhAzWpZHSiYPc09/NwDQIGyg==" crossorigin="anonymous"></script><script src="https://cdnjs.cloudflare.com/ajax/libs/mark.js/8.11.1/mark.min.js" integrity="sha512-5CYOlHXGh6QpOFA/TeTylKLWfB3ftPsde7AnmhuitiTX4K5SqCLBeKro6sPS8ilsz1Q4NRx3v8Ko2IBiszzdww==" crossorigin="anonymous"></script><!-- pkgdown --><script src="../pkgdown.js"></script><meta property="og:title" content="Using Tensorboard">
<meta property="og:description" content="deepG">
<meta property="og:image" content="https://genomenet.github.io/deepG/logo.png">
<!-- mathjax --><script src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.5/MathJax.js" integrity="sha256-nvJJv9wWKEm88qvoQl9ekL2J+k/RWIsaSScxxlsrv8k=" crossorigin="anonymous"></script><script src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.5/config/TeX-AMS-MML_HTMLorMML.js" integrity="sha256-84DKXVJXs0/F8OTMzX4UR909+jtl4G7SPypPavF+GfA=" crossorigin="anonymous"></script><!--[if lt IE 9]>
<script src="https://oss.maxcdn.com/html5shiv/3.7.3/html5shiv.min.js"></script>
<script src="https://oss.maxcdn.com/respond/1.4.2/respond.min.js"></script>
<![endif]-->
</head>
<body>
    <a href="#main" class="visually-hidden-focusable">Skip to contents</a>
    

    <nav class="navbar fixed-top navbar-light navbar-expand-lg bg-light" data-bs-theme="light"><div class="container">
    
    <a class="navbar-brand me-2" href="../index.html">deepG</a>

    <small class="nav-text text-default me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="Released version">0.3.0</small>

    
    <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
      <span class="navbar-toggler-icon"></span>
    </button>

    <div id="navbar" class="collapse navbar-collapse ms-3">
      <ul class="navbar-nav me-auto">
<li class="nav-item">
  <a class="nav-link" href="../reference/index.html">
    <span class="fa fa fa fa-file-alt"></span>
     
    Reference
  </a>
</li>
<li class="nav-item dropdown">
  <a href="#" class="nav-link dropdown-toggle" data-bs-toggle="dropdown" role="button" aria-expanded="false" aria-haspopup="true" id="dropdown-notebooks">Notebooks</a>
  <div class="dropdown-menu" aria-labelledby="dropdown-notebooks">
    <a class="external-link dropdown-item" href="https://colab.research.google.com/drive/175jIdXcDcgPUvaBo2rH2Lupbpjnp5O7G?usp=sharing">deepG tutorial</a>
    <a class="external-link dropdown-item" href="https://colab.research.google.com/drive/1Eolc0koMNM1zkuO4XyVM58ImeF1BpRiH?usp=sharing">Read-length level: Human contamination</a>
    <a class="external-link dropdown-item" href="https://colab.research.google.com/drive/1yiXSwFafXpMLHaov9iBTQLIDZ6bK1zYX?usp=sharing">Locus level: CRISPR detection</a>
    <a class="external-link dropdown-item" href="https://colab.research.google.com/drive/1G7bOFEX87cZNrM2tdRtTdkrZn5fM__g0?usp=sharing">Gene level: 16S rRNA detection</a>
    <a class="external-link dropdown-item" href="https://colab.research.google.com/drive/1BCggL-tfQF136YeJ8cKKi-zoBEDMgkNh?usp=sharing">Genome level: Bacterial morphology (Sporulation)</a>
    <a class="external-link dropdown-item" href="https://colab.research.google.com/drive/10xpRzGd3JeBAbqQYSCxzQUMctt01sx9D?usp=sharing">Full metagenome level: Colorectal cancer prediction</a>
    <a class="external-link dropdown-item" href="https://colab.research.google.com/drive/1kyYK7IU7GSfdpDzO_a8U3_qD4i3zTu6w?usp=sharing">BERT with deepG</a>
  </div>
</li>
<li class="active nav-item dropdown">
  <a href="#" class="nav-link dropdown-toggle" data-bs-toggle="dropdown" role="button" aria-expanded="false" aria-haspopup="true" id="dropdown-tutorials">Tutorials</a>
  <div class="dropdown-menu" aria-labelledby="dropdown-tutorials">
    <a class="dropdown-item" href="../articles/getting_started.html">Getting Started</a>
    <a class="dropdown-item" href="../articles/training_types.html">Training types</a>
    <a class="dropdown-item" href="../articles/data_generator.html">Data generator</a>
    <a class="dropdown-item" href="../articles/using_tb.html">Using tensorboard</a>
    <a class="dropdown-item" href="../articles/integrated_gradient.html">Integrated Gradient</a>
  </div>
</li>
      </ul>
<form class="form-inline my-2 my-lg-0" role="search">
        <input type="search" class="form-control me-sm-2" aria-label="Toggle navigation" name="search-input" data-search-index="../search.json" id="search-input" placeholder="Search for" autocomplete="off">
</form>

      <ul class="navbar-nav">
<li class="nav-item">
  <a class="external-link nav-link" href="https://github.com/GenomeNet/deepG/" aria-label="github">
    <span class="fab fa fab fa-github fa-lg"></span>
     
  </a>
</li>
      </ul>
</div>

    
  </div>
</nav><div class="container template-article">




<div class="row">
  <main id="main" class="col-md-9"><div class="page-header">
      <img src="../logo.png" class="logo" alt=""><h1>Using Tensorboard</h1>
            
      
      <small class="dont-index">Source: <a href="https://github.com/GenomeNet/deepG/blob/HEAD/vignettes/using_tb.Rmd" class="external-link"><code>vignettes/using_tb.Rmd</code></a></small>
      <div class="d-none name"><code>using_tb.Rmd</code></div>
    </div>

    
    
<div class="sourceCode" id="cb1"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span><span class="co"># devtools::install_github("GenomeNet/deepG")</span></span>
<span><span class="kw"><a href="https://rdrr.io/r/base/library.html" class="external-link">library</a></span><span class="op">(</span><span class="va"><a href="https://github.com/GenomeNet/deepG" class="external-link">deepG</a></span><span class="op">)</span></span>
<span><span class="kw"><a href="https://rdrr.io/r/base/library.html" class="external-link">library</a></span><span class="op">(</span><span class="va"><a href="https://magrittr.tidyverse.org" class="external-link">magrittr</a></span><span class="op">)</span></span></code></pre></div>
<style type="text/css">
mark.in {
  background-color: CornflowerBlue;
}

mark.out {
  background-color: IndianRed;
}

</style>
<p>Tensorflow offers the
<a href="https://www.tensorflow.org/tensorboard" class="external-link">Tensorboard</a>
application to visualize the training process of our networks. DeepG
expands on some of the default Tensorboard options and implements some
custom settings. We train again a model that can differentiate sequences
based on the GC content, as described in the
<a href="getting_started.html">Getting started tutorial</a>.</p>
<p>We start by creating our data. To show the difference between
accuracy and balanced accuracy, we create 3 times more data with high GC
content.</p>
<div class="sourceCode" id="cb2"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span><span class="fu"><a href="https://rdrr.io/r/base/Random.html" class="external-link">set.seed</a></span><span class="op">(</span><span class="fl">123</span><span class="op">)</span></span>
<span><span class="va">vocabulary</span> <span class="op">&lt;-</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="st">"A"</span>, <span class="st">"C"</span>, <span class="st">"G"</span>, <span class="st">"T"</span><span class="op">)</span></span>
<span></span>
<span><span class="va">data_type</span> <span class="op">&lt;-</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="st">"train"</span>, <span class="st">"validation"</span><span class="op">)</span></span>
<span></span>
<span><span class="kw">for</span> <span class="op">(</span><span class="va">i</span> <span class="kw">in</span> <span class="fl">1</span><span class="op">:</span><span class="fu"><a href="https://rdrr.io/r/base/length.html" class="external-link">length</a></span><span class="op">(</span><span class="va">data_type</span><span class="op">)</span><span class="op">)</span> <span class="op">{</span></span>
<span>  </span>
<span>  <span class="va">temp_file</span> <span class="op">&lt;-</span> <span class="fu"><a href="https://rdrr.io/r/base/tempfile.html" class="external-link">tempfile</a></span><span class="op">(</span><span class="op">)</span></span>
<span>  <span class="fu"><a href="https://rdrr.io/r/base/assign.html" class="external-link">assign</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/r/base/paste.html" class="external-link">paste0</a></span><span class="op">(</span><span class="va">data_type</span><span class="op">[</span><span class="va">i</span><span class="op">]</span>, <span class="st">"_dir"</span><span class="op">)</span>, <span class="va">temp_file</span><span class="op">)</span></span>
<span>  <span class="fu"><a href="https://rdrr.io/r/base/files2.html" class="external-link">dir.create</a></span><span class="op">(</span><span class="va">temp_file</span><span class="op">)</span></span>
<span>  </span>
<span>  <span class="kw">for</span> <span class="op">(</span><span class="va">j</span> <span class="kw">in</span> <span class="fl">1</span><span class="op">:</span><span class="fl">6</span><span class="op">)</span> <span class="op">{</span></span>
<span>    </span>
<span>    <span class="kw">if</span> <span class="op">(</span><span class="va">j</span> <span class="op"><a href="https://rdrr.io/r/base/Arithmetic.html" class="external-link">%%</a></span> <span class="fl">2</span> <span class="op">==</span> <span class="fl">1</span><span class="op">)</span> <span class="op">{</span></span>
<span>      <span class="va">header</span> <span class="op">&lt;-</span> <span class="st">"high_gc"</span></span>
<span>      <span class="va">prob</span> <span class="op">&lt;-</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="fl">0.1</span>, <span class="fl">0.4</span>, <span class="fl">0.4</span>, <span class="fl">0.1</span><span class="op">)</span></span>
<span>    <span class="op">}</span> <span class="kw">else</span> <span class="op">{</span></span>
<span>      <span class="va">header</span> <span class="op">&lt;-</span> <span class="st">"equal_dist"</span></span>
<span>      <span class="va">prob</span> <span class="op">&lt;-</span> <span class="fu"><a href="https://rdrr.io/r/base/rep.html" class="external-link">rep</a></span><span class="op">(</span><span class="fl">0.25</span>, <span class="fl">4</span><span class="op">)</span></span>
<span>    <span class="op">}</span></span>
<span>    <span class="va">fasta_name_start</span> <span class="op">&lt;-</span> <span class="fu"><a href="https://rdrr.io/r/base/paste.html" class="external-link">paste0</a></span><span class="op">(</span><span class="va">header</span>, <span class="st">"_"</span>, <span class="va">data_type</span><span class="op">[</span><span class="va">i</span><span class="op">]</span><span class="op">)</span></span>
<span>    </span>
<span>    <span class="fu"><a href="../reference/create_dummy_data.html">create_dummy_data</a></span><span class="op">(</span>file_path <span class="op">=</span> <span class="va">temp_file</span>,</span>
<span>                      num_files <span class="op">=</span> <span class="fl">2</span>,</span>
<span>                      seq_length <span class="op">=</span> <span class="fl">100</span>, </span>
<span>                      num_seq <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/ifelse.html" class="external-link">ifelse</a></span><span class="op">(</span><span class="va">j</span> <span class="op"><a href="https://rdrr.io/r/base/Arithmetic.html" class="external-link">%%</a></span> <span class="fl">2</span> <span class="op">==</span> <span class="fl">1</span>, <span class="fl">6</span>, <span class="fl">2</span><span class="op">)</span>, <span class="co"># create more sequences for high GC content</span></span>
<span>                      header <span class="op">=</span> <span class="va">header</span>,</span>
<span>                      prob <span class="op">=</span> <span class="va">prob</span>,</span>
<span>                      fasta_name_start <span class="op">=</span> <span class="va">fasta_name_start</span>,</span>
<span>                      vocabulary <span class="op">=</span> <span class="va">vocabulary</span><span class="op">)</span></span>
<span>  <span class="op">}</span></span>
<span>  </span>
<span><span class="op">}</span></span></code></pre></div>
<div class="sourceCode" id="cb3"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span><span class="fu"><a href="https://rdrr.io/r/base/list.files.html" class="external-link">list.files</a></span><span class="op">(</span><span class="va">train_dir</span><span class="op">)</span></span></code></pre></div>
<pre><code><span><span class="co">## [1] "equal_dist_train_1.fasta" "equal_dist_train_2.fasta"</span></span>
<span><span class="co">## [3] "high_gc_train_1.fasta"    "high_gc_train_2.fasta"</span></span></code></pre>
<div class="sourceCode" id="cb5"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span><span class="fu"><a href="https://rdrr.io/r/base/list.files.html" class="external-link">list.files</a></span><span class="op">(</span><span class="va">validation_dir</span><span class="op">)</span></span></code></pre></div>
<pre><code><span><span class="co">## [1] "equal_dist_validation_1.fasta" "equal_dist_validation_2.fasta"</span></span>
<span><span class="co">## [3] "high_gc_validation_1.fasta"    "high_gc_validation_2.fasta"</span></span></code></pre>
<p>To use tensorboard, we first need to create a folder where we store
our tensorboard logs.</p>
<div class="sourceCode" id="cb7"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span><span class="co"># create folder for tensorboard logs</span></span>
<span><span class="va">tb_dir</span> <span class="op">&lt;-</span> <span class="fu"><a href="https://rdrr.io/r/base/tempfile.html" class="external-link">tempfile</a></span><span class="op">(</span><span class="op">)</span></span>
<span><span class="fu"><a href="https://rdrr.io/r/base/files2.html" class="external-link">dir.create</a></span><span class="op">(</span><span class="va">tb_dir</span><span class="op">)</span></span></code></pre></div>
<p>When creating our model, we can add some additional metrics to
observe like AUC, F1 and balanced accuracy.</p>
<div class="sourceCode" id="cb8"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span><span class="va">maxlen</span> <span class="op">&lt;-</span> <span class="fl">50</span></span>
<span><span class="va">model</span> <span class="op">&lt;-</span> <span class="fu"><a href="../reference/create_model_lstm_cnn.html">create_model_lstm_cnn</a></span><span class="op">(</span>maxlen <span class="op">=</span> <span class="va">maxlen</span>,</span>
<span>                               filters <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="fl">8</span><span class="op">)</span>,</span>
<span>                               kernel_size <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="fl">12</span><span class="op">)</span>,</span>
<span>                               pool_size <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="fl">3</span><span class="op">)</span>,</span>
<span>                               layer_lstm <span class="op">=</span> <span class="fl">8</span>,</span>
<span>                               auc_metric <span class="op">=</span> <span class="cn">TRUE</span>,</span>
<span>                               f1_metric <span class="op">=</span> <span class="cn">TRUE</span>,</span>
<span>                               bal_acc <span class="op">=</span> <span class="cn">TRUE</span>,</span>
<span>                               layer_dense <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="fl">4</span>, <span class="fl">2</span><span class="op">)</span>,</span>
<span>                               model_seed <span class="op">=</span> <span class="fl">3</span><span class="op">)</span></span></code></pre></div>
<pre><code><span><span class="co">## Will flatten y_true and y_pred matrices to a single vector for evaluation</span></span>
<span><span class="co">##             rather than computing separate F1 scores for each class and taking the mean.</span></span></code></pre>
<pre><code><span><span class="co">## Model: "model"</span></span>
<span><span class="co">## _________________________________________________________________</span></span>
<span><span class="co">##  Layer (type)                Output Shape              Param #   </span></span>
<span><span class="co">## =================================================================</span></span>
<span><span class="co">##  input_1 (InputLayer)        [(None, 50, 4)]           0         </span></span>
<span><span class="co">##                                                                  </span></span>
<span><span class="co">##  conv1d (Conv1D)             (None, 50, 8)             392       </span></span>
<span><span class="co">##                                                                  </span></span>
<span><span class="co">##  max_pooling1d (MaxPooling1  (None, 16, 8)             0         </span></span>
<span><span class="co">##  D)                                                              </span></span>
<span><span class="co">##                                                                  </span></span>
<span><span class="co">##  batch_normalization (Batch  (None, 16, 8)             32        </span></span>
<span><span class="co">##  Normalization)                                                  </span></span>
<span><span class="co">##                                                                  </span></span>
<span><span class="co">##  lstm (LSTM)                 (None, 8)                 544       </span></span>
<span><span class="co">##                                                                  </span></span>
<span><span class="co">##  dense (Dense)               (None, 4)                 36        </span></span>
<span><span class="co">##                                                                  </span></span>
<span><span class="co">##  dense_1 (Dense)             (None, 2)                 10        </span></span>
<span><span class="co">##                                                                  </span></span>
<span><span class="co">## =================================================================</span></span>
<span><span class="co">## Total params: 1014 (3.96 KB)</span></span>
<span><span class="co">## Trainable params: 998 (3.90 KB)</span></span>
<span><span class="co">## Non-trainable params: 16 (64.00 Byte)</span></span>
<span><span class="co">## _________________________________________________________________</span></span></code></pre>
<p>Finally we can train the model.</p>
<div class="sourceCode" id="cb11"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span><span class="va">hist</span> <span class="op">&lt;-</span> <span class="fu"><a href="../reference/train_model.html">train_model</a></span><span class="op">(</span><span class="va">model</span>,</span>
<span>                    train_type <span class="op">=</span> <span class="st">"label_header"</span>,</span>
<span>                    run_name <span class="op">=</span> <span class="st">"gc_model_1"</span>,</span>
<span>                    path <span class="op">=</span> <span class="va">train_dir</span>,</span>
<span>                    path_val <span class="op">=</span> <span class="va">validation_dir</span>,</span>
<span>                    epochs <span class="op">=</span> <span class="fl">5</span>,</span>
<span>                    steps_per_epoch <span class="op">=</span> <span class="fl">2</span>, <span class="co"># 20</span></span>
<span>                    batch_size <span class="op">=</span> <span class="fl">64</span>,</span>
<span>                    step <span class="op">=</span> <span class="fl">50</span>, </span>
<span>                    path_tensorboard <span class="op">=</span> <span class="va">tb_dir</span>, <span class="co"># path to tensorboard logs</span></span>
<span>                    tb_images <span class="op">=</span> <span class="cn">F</span>, <span class="co"># show confusion matrix in tensorboard</span></span>
<span>                    vocabulary_label <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="st">"high_gc"</span>, <span class="st">"equal_dist"</span><span class="op">)</span><span class="op">)</span></span>
<span>                    </span>
<span><span class="fu"><a href="https://rdrr.io/r/graphics/plot.default.html" class="external-link">plot</a></span><span class="op">(</span><span class="va">hist</span><span class="op">)</span></span></code></pre></div>
<p>Use the following command to open tensorboard in a browser:</p>
<div class="sourceCode" id="cb12"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span><span class="fu">keras</span><span class="fu">::</span><span class="fu"><a href="https://rdrr.io/pkg/tensorflow/man/tensorboard.html" class="external-link">tensorboard</a></span><span class="op">(</span><span class="va">tb_dir</span><span class="op">)</span></span></code></pre></div>
<p>We can observe the scores for loss, accuracy, balanced accuracy, F1
and AUC under the “SCALARS” tab.</p>
<p><img src="tb_images/loss.png" height="300"></p>
<p><img src="tb_images/acc.png" height="300"></p>
<p><img src="tb_images/bal_acc.png" height="300"></p>
<p><img src="tb_images/f1.png"></p>
<p><img src="tb_images/auc.png" height="300"></p>
<p>We can also observe how the learning rate might have changed</p>
<p><img src="tb_images/lr.png" height="300"></p>
<p>In the “training files seen” window, we can observe how often we
iterated over the training files.</p>
<p><img src="tb_images/files_seen.png" height="300"></p>
<p>In the “IMAGES” tab we can see a confusion matrix for the train and
validation scores for every epoch</p>
<p><img src="tb_images/cm_train.png" height="300"></p>
<p><img src="tb_images/cm_val.png" height="300"></p>
<p>In the “HPARAM” tab you can see hyper parameters of each run</p>
<p><img src="tb_images/hparam.png" height="300"></p>
  </main>
</div>



    <footer><div class="pkgdown-footer-left">
  <p>Developed by Philipp Münch, René Mreches, Martin Binder, Hüseyin Anil Gündüz, Xiao-Yin To, Alice McHardy.</p>
</div>

<div class="pkgdown-footer-right">
  <p>Site built with <a href="https://pkgdown.r-lib.org/" class="external-link">pkgdown</a> 2.0.9.</p>
</div>

    </footer>
</div>

  

  

  </body>
</html>