save_model.html



<!DOCTYPE html>
<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
<head>
  <meta charset="utf-8">
  
  <meta name="viewport" content="width=device-width, initial-scale=1.0">
  
  <title>Design Doc: Save Model &mdash; PaddlePaddle  documentation</title>
  

    <link rel="stylesheet" href="../../_static/css/theme.css" type="text/css" />
  

        <link rel="index" title="Index"
              href="../../genindex.html"/>
        <link rel="search" title="Search" href="../../search.html"/>
    <link rel="top" title="PaddlePaddle  documentation" href="../../index.html"/> 
<script>
var _hmt = _hmt || [];
(function() {
  var hm = document.createElement("script");
  hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
  var s = document.getElementsByTagName("script")[0]; 
  s.parentNode.insertBefore(hm, s);
})();
</script>


  <script src="../../_static/js/modernizr.min.js"></script>

</head>

<body class="wy-body-for-nav" role="document">

  <div class="wy-grid-for-nav">

    
    <nav data-toggle="wy-nav-shift" class="wy-nav-side">
      <div class="wy-side-scroll">
        <div class="wy-side-nav-search">
          

            <a href="../../index_en.html" class="icon icon-home"> PaddlePaddle
          

          </a>

          
<div role="search">
  <form id="rtd-search-form" class="wy-form" action="../../search.html" method="get">
    <input type="text" name="q" placeholder="Search docs" />
    <input type="hidden" name="check_keywords" value="yes" />
    <input type="hidden" name="area" value="default" />
  </form>
</div>

          
        </div>

        <div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="main navigation">
          
<nav class="doc-menu-vertical" role="navigation">

<ul>
<li class="toctree-l1"><a class="reference internal" href="../../getstarted/index_en.html">GET STARTED</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../../getstarted/quickstart_en.html">Quick Start</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../getstarted/concepts/use_concepts_en.html">Basic Concept</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../build_and_install/index_en.html">Install and Build</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../../build_and_install/pip_install_en.html">Install using pip</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../build_and_install/docker_install_en.html">Run in Docker Containers</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../build_and_install/build_from_source_en.html">Build from Sources</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../howto/index_en.html">HOW TO</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../../howto/cmd_parameter/index_en.html">Set Command-line Parameters</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../../howto/cmd_parameter/use_case_en.html">Use Case</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../howto/cmd_parameter/arguments_en.html">Argument Outline</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../howto/cmd_parameter/detail_introduction_en.html">Detail Description</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../../howto/cluster/index_en.html">Distributed Training</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../../howto/cluster/preparations_en.html">Preparations</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../howto/cluster/cmd_argument_en.html">Command-line arguments</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../howto/cluster/multi_cluster/index_en.html">Use different clusters</a><ul>
<li class="toctree-l4"><a class="reference internal" href="../../howto/cluster/multi_cluster/fabric_en.html">Fabric</a></li>
<li class="toctree-l4"><a class="reference internal" href="../../howto/cluster/multi_cluster/openmpi_en.html">OpenMPI</a></li>
<li class="toctree-l4"><a class="reference internal" href="../../howto/cluster/multi_cluster/k8s_en.html">Kubernetes</a></li>
<li class="toctree-l4"><a class="reference internal" href="../../howto/cluster/multi_cluster/k8s_aws_en.html">Kubernetes on AWS</a></li>
</ul>
</li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../../howto/rnn/index_en.html">RNN Models</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../../howto/rnn/rnn_config_en.html">RNN Configuration</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../howto/rnn/recurrent_group_en.html">Recurrent Group Tutorial</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../howto/rnn/hierarchical_layer_en.html">Layers supporting hierarchical sequence as input</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../howto/rnn/hrnn_rnn_api_compare_en.html">API comparision between RNN and hierarchical RNN</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../../howto/optimization/gpu_profiling_en.html">Tune GPU Performance</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../dev/index_en.html">Development</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../../dev/contribute_to_paddle_en.html">Contribute Code</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../dev/write_docs_en.html">Contribute Documentation</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../faq/index_en.html">FAQ</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../../faq/build_and_install/index_en.html">Install, Build and Unit test</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../faq/model/index_en.html">Model Configuration</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../faq/parameter/index_en.html">Parameter Setting</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../faq/local/index_en.html">Local Training and Prediction</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../faq/cluster/index_en.html">Cluster Training and Prediction</a></li>
</ul>
</li>
</ul>

</nav>

        </div>
      </div>
    </nav>

    <section data-toggle="wy-nav-shift" class="wy-nav-content-wrap">

      
      <nav class="wy-nav-top" role="navigation" aria-label="top navigation">
        <i data-toggle="wy-nav-top" class="fa fa-bars"></i>
        <a href="../../index_en.html">PaddlePaddle</a>
      </nav>


      <div class="wy-nav-content">
        <div class="rst-content">
          

<div role="navigation" aria-label="breadcrumbs navigation">
  <ul class="wy-breadcrumbs">
    <li><a href="../../index_en.html">Docs</a> &raquo;</li>
      
    <li>Design Doc: Save Model</li>
      <li class="wy-breadcrumbs-aside">
        
          
            <a href="../../_sources/design/cluster_train/save_model.md.txt" rel="nofollow"> View page source</a>
          
        
      </li>
  </ul>
  <hr/>
</div>
          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
           <div itemprop="articleBody">
            
  <div class="section" id="design-doc-save-model">
<span id="design-doc-save-model"></span><h1>Design Doc: Save Model<a class="headerlink" href="#design-doc-save-model" title="Permalink to this headline">¶</a></h1>
<div class="section" id="overview">
<span id="overview"></span><h2>Overview<a class="headerlink" href="#overview" title="Permalink to this headline">¶</a></h2>
<p>The model is the output of the training process. There are two
ways from which user can obtain a model:</p>
<ul class="simple">
<li>Save model triggered by user code: user code asks PaddlePaddle to
save a model.</li>
<li>Convert model from the checkpoint: model being converted from
pservers&#8217; periodic checkpoint. In this way, the user can cancel a
job at any time, and still have a relatively fresh model (we
checkpoint around every 5 minutes).</li>
</ul>
<div class="section" id="trainer-saving-model-vs-pservers-saving-model">
<span id="trainer-saving-model-vs-pservers-saving-model"></span><h3>Trainer Saving Model vs. Pservers Saving Model<a class="headerlink" href="#trainer-saving-model-vs-pservers-saving-model" title="Permalink to this headline">¶</a></h3>
<p>Both trainers and pservers have access to the model. So the model can
be saved from a trainer or pservers. We need to decide where the model
is saved from.</p>
<div class="section" id="dense-update-vs-sparse-update">
<span id="dense-update-vs-sparse-update"></span><h4>Dense Update vs. Sparse Update<a class="headerlink" href="#dense-update-vs-sparse-update" title="Permalink to this headline">¶</a></h4>
<p>There are two types of model update methods: dense update and sparse
update (when the model parameter is configured to be sparse).</p>
<ul>
<li><p class="first">Dense update</p>
<p>Every trainer has it&#8217;s own full copy of the model. Every model
update will update the entire model.</p>
</li>
<li><p class="first">Sparse update</p>
<p>The training input is sparse, and the trainer does not have the
entire model. It will only download the sub-model necessary related
to the input. When updating the model, only the sub-model related to
the training input is updated.</p>
</li>
</ul>
</div>
<div class="section" id="pservers-saving-model">
<span id="pservers-saving-model"></span><h4>Pservers Saving Model<a class="headerlink" href="#pservers-saving-model" title="Permalink to this headline">¶</a></h4>
<p>The benefit of letting pservers save model is they have the entire
model all the time. However, since pservers are on different nodes, it
requires a merging process to merge model shards into the same
model. Thus requires the pservers to write models to a distributed
filesystem, making the checkpoint shards visible to the merge program.</p>
</div>
<div class="section" id="trainer-saving-model">
<span id="trainer-saving-model"></span><h4>Trainer Saving Model<a class="headerlink" href="#trainer-saving-model" title="Permalink to this headline">¶</a></h4>
<p>The benefit of letting one trainer to save the model is it does not
require a distributed filesystem. And it&#8217;s reusing the same save model
logic when training locally - except when doing sparse update, the
trainer needs to download the entire model during the saving process.</p>
</div>
<div class="section" id="conclusion">
<span id="conclusion"></span><h4>Conclusion<a class="headerlink" href="#conclusion" title="Permalink to this headline">¶</a></h4>
<p>Given trainer saving model does not require a distributed filesystem,
and is an intuitive extension to trainer saving model when training
locally, we decide to let the trainer save the model when doing
distributed training.</p>
</div>
</div>
<div class="section" id="convert-model-from-checkpoint">
<span id="convert-model-from-checkpoint"></span><h3>Convert Model from Checkpoint<a class="headerlink" href="#convert-model-from-checkpoint" title="Permalink to this headline">¶</a></h3>
<p>TODO</p>
</div>
</div>
<div class="section" id="timeline">
<span id="timeline"></span><h2>Timeline<a class="headerlink" href="#timeline" title="Permalink to this headline">¶</a></h2>
<p>We first implement trainer save the model. Converting the latest
snapshot to a model will be a TODO for future.</p>
</div>
<div class="section" id="trainer-save-model">
<span id="trainer-save-model"></span><h2>Trainer Save Model<a class="headerlink" href="#trainer-save-model" title="Permalink to this headline">¶</a></h2>
<div class="section" id="trainer-election">
<span id="trainer-election"></span><h3>Trainer Election<a class="headerlink" href="#trainer-election" title="Permalink to this headline">¶</a></h3>
<p>One trainer will be elected as the one to save the model. When using
etcd, trainer ID is a randomly generated UUID, the trainer will
contact the master server requesting to save the model, and find out
if itself is elected. When the master server is not used, unique
trainer IDs will be given by the administrator, the trainer whose ID
is &#8220;0&#8221; is elected to save the model.</p>
</div>
<div class="section" id="model-save-path">
<span id="model-save-path"></span><h3>Model Save Path<a class="headerlink" href="#model-save-path" title="Permalink to this headline">¶</a></h3>
<p>Each trainer will be given the directory to save the model. The
elected trainer will save the model to
<code class="docutils literal"><span class="pre">given-directory/trainerID</span></code>. Since the trainer ID is unique, this
would prevent concurrent save to the same file when multiple trainers
are elected to save the model when split-brain problem happens.</p>
</div>
<div class="section" id="what-happens-when-model-is-saving">
<span id="what-happens-when-model-is-saving"></span><h3>What Happens When Model Is Saving<a class="headerlink" href="#what-happens-when-model-is-saving" title="Permalink to this headline">¶</a></h3>
<p>It takes some time to save model, we need to define what will happen
when save model is taking place.</p>
<p>When doing dense update, the trainer uses the local model. Pservers
does not need to pause model update.</p>
<p>When doing sparse update. The trainer needs to download the entire
model while saving. To get the most accurate model, the model update
needs to be paused before the download starts and resumed after the
download finishes. Otherwise, the trainer gets a model that is
&#8220;polluted&#8221;: some part of the model is old, some part of the model is
new.</p>
<p>It&#8217;s unclear that the &#8220;polluted&#8221; model will be inferior due to the
stochastic nature of deep learning, and pausing the model update will
add more complexity to the system. Since supporting sparse update is a
TODO item. We defer the evaluation of pause the model update or not
during saving model to the future.</p>
</div>
</div>
</div>


           </div>
          </div>
          <footer>
  

  <hr/>

  <div role="contentinfo">
    <p>
        &copy; Copyright 2016, PaddlePaddle developers.

    </p>
  </div>
  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 

</footer>

        </div>
      </div>

    </section>

  </div>
  

    <script type="text/javascript">
        var DOCUMENTATION_OPTIONS = {
            URL_ROOT:'../../',
            VERSION:'',
            COLLAPSE_INDEX:false,
            FILE_SUFFIX:'.html',
            HAS_SOURCE:  true
        };
    </script>
      <script type="text/javascript" src="../../_static/jquery.js"></script>
      <script type="text/javascript" src="../../_static/underscore.js"></script>
      <script type="text/javascript" src="../../_static/doctools.js"></script>
      <script type="text/javascript" src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.0/MathJax.js?config=TeX-AMS-MML_HTMLorMML"></script>

  
    <script type="text/javascript" src="../../_static/js/theme.js"></script>
  

  <script type="text/javascript">
      jQuery(function () {
          SphinxRtdTheme.StickyNav.enable();
      });
  </script>
   

</body>
</html>