parallel_do.html 26.6 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219


<!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: Parallel_Do in PaddlePaddle &mdash; PaddlePaddle  文档</title>
  

  
  

  

  
  
    

  

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

  
  
        <link rel="index" title="索引"
              href="../genindex.html"/>
        <link rel="search" title="搜索" href="../search.html"/>
    <link rel="top" title="PaddlePaddle  文档" href="../index.html"/> 

  <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
  <link rel="stylesheet" href="../_static/css/override.css" type="text/css" />
  <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">

  
  <header class="site-header">
    <div class="site-logo">
      <a href="/"><img src="../_static/images/PP_w.png"></a>
    </div>
    <div class="site-nav-links">
      <div class="site-menu">
        <a class="fork-on-github" href="https://github.com/PaddlePaddle/Paddle" target="_blank"><i class="fa fa-github"></i>Fork me on Github</a>
        <div class="language-switcher dropdown">
          <a type="button" data-toggle="dropdown">
            <span>English</span>
            <i class="fa fa-angle-up"></i>
            <i class="fa fa-angle-down"></i>
          </a>
          <ul class="dropdown-menu">
            <li><a href="/doc_cn">中文</a></li>
            <li><a href="/doc">English</a></li>
          </ul>
        </div>
        <ul class="site-page-links">
          <li><a href="/">Home</a></li>
        </ul>
      </div>
      <div class="doc-module">
        
        <ul>
<li class="toctree-l1"><a class="reference internal" href="../getstarted/index_cn.html">新手入门</a></li>
<li class="toctree-l1"><a class="reference internal" href="../build_and_install/index_cn.html">安装与编译</a></li>
<li class="toctree-l1"><a class="reference internal" href="../howto/index_cn.html">进阶使用</a></li>
<li class="toctree-l1"><a class="reference internal" href="../dev/index_cn.html">开发标准</a></li>
<li class="toctree-l1"><a class="reference internal" href="../faq/index_cn.html">FAQ</a></li>
</ul>

        
<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>
  </header>
  
  <div class="main-content-wrap">

    
    <nav class="doc-menu-vertical" role="navigation">
        
          
          <ul>
<li class="toctree-l1"><a class="reference internal" href="../getstarted/index_cn.html">新手入门</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../getstarted/quickstart_cn.html">快速开始</a></li>
<li class="toctree-l2"><a class="reference internal" href="../getstarted/concepts/use_concepts_cn.html">基本使用概念</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../build_and_install/index_cn.html">安装与编译</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../build_and_install/pip_install_cn.html">使用pip安装</a></li>
<li class="toctree-l2"><a class="reference internal" href="../build_and_install/docker_install_cn.html">使用Docker安装运行</a></li>
<li class="toctree-l2"><a class="reference internal" href="../build_and_install/build_from_source_cn.html">从源码编译</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../howto/index_cn.html">进阶使用</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../howto/cmd_parameter/index_cn.html">命令行参数设置</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../howto/cmd_parameter/use_case_cn.html">使用案例</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/cmd_parameter/arguments_cn.html">参数概述</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/cmd_parameter/detail_introduction_cn.html">细节描述</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../howto/cluster/index_cn.html">分布式训练</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../howto/cluster/preparations_cn.html">环境准备</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/cluster/cmd_argument_cn.html">启动参数说明</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/cluster/multi_cluster/index_cn.html">在不同集群中运行</a><ul>
<li class="toctree-l4"><a class="reference internal" href="../howto/cluster/multi_cluster/fabric_cn.html">使用fabric启动集群训练</a></li>
<li class="toctree-l4"><a class="reference internal" href="../howto/cluster/multi_cluster/openmpi_cn.html">在OpenMPI集群中提交训练作业</a></li>
<li class="toctree-l4"><a class="reference internal" href="../howto/cluster/multi_cluster/k8s_cn.html">Kubernetes单机训练</a></li>
<li class="toctree-l4"><a class="reference internal" href="../howto/cluster/multi_cluster/k8s_distributed_cn.html">Kubernetes分布式训练</a></li>
<li class="toctree-l4"><a class="reference internal" href="../howto/cluster/multi_cluster/k8s_aws_cn.html">Distributed PaddlePaddle Training on AWS with Kubernetes</a></li>
</ul>
</li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../howto/capi/index_cn.html">C-API预测库</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../howto/capi/compile_paddle_lib_cn.html">安装与编译C-API预测库</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/capi/organization_of_the_inputs_cn.html">输入/输出数据组织</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/capi/workflow_of_capi_cn.html">C-API使用流程</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../howto/rnn/index_cn.html">RNN模型</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../howto/rnn/rnn_config_cn.html">RNN配置</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/rnn/recurrent_group_cn.html">Recurrent Group教程</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/rnn/hierarchical_layer_cn.html">支持双层序列作为输入的Layer</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/rnn/hrnn_rnn_api_compare_cn.html">单双层RNN API对比介绍</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../howto/optimization/gpu_profiling_cn.html">GPU性能调优</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../dev/index_cn.html">开发标准</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../dev/contribute_to_paddle_cn.html">如何贡献代码</a></li>
<li class="toctree-l2"><a class="reference internal" href="../dev/write_docs_cn.html">如何贡献文档</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../faq/index_cn.html">FAQ</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../faq/build_and_install/index_cn.html">编译安装与单元测试</a></li>
<li class="toctree-l2"><a class="reference internal" href="../faq/model/index_cn.html">模型配置</a></li>
<li class="toctree-l2"><a class="reference internal" href="../faq/parameter/index_cn.html">参数设置</a></li>
<li class="toctree-l2"><a class="reference internal" href="../faq/local/index_cn.html">本地训练与预测</a></li>
<li class="toctree-l2"><a class="reference internal" href="../faq/cluster/index_cn.html">集群训练与预测</a></li>
</ul>
</li>
</ul>

        
    </nav>
    
    <section class="doc-content-wrap">

      

 







<div role="navigation" aria-label="breadcrumbs navigation">
  <ul class="wy-breadcrumbs">
      
    <li>Design Doc: Parallel_Do in PaddlePaddle</li>
  </ul>
</div>
      
      <div class="wy-nav-content" id="doc-content">
        <div class="rst-content">
          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
           <div itemprop="articleBody">
            
  <div class="section" id="design-doc-parallel-do-in-paddlepaddle">
<span id="design-doc-parallel-do-in-paddlepaddle"></span><h1>Design Doc: Parallel_Do in PaddlePaddle<a class="headerlink" href="#design-doc-parallel-do-in-paddlepaddle" title="永久链接至标题"></a></h1>
<p>In PaddlePaddle, we use parallel_do primitive to represent multithread data parallel processing.</p>
<div class="section" id="design-overview">
<span id="design-overview"></span><h2>Design overview<a class="headerlink" href="#design-overview" title="永久链接至标题"></a></h2>
<p>The definition of a parallel_do op looks like the following</p>
<div class="highlight-c++"><div class="highlight"><pre><span></span><span class="n">AddInput</span><span class="p">(</span><span class="n">kInputs</span><span class="p">,</span> <span class="s">&quot;Inputs needed to be split onto different devices&quot;</span><span class="p">).</span><span class="n">AsDuplicable</span><span class="p">();</span>
<span class="n">AddInput</span><span class="p">(</span><span class="n">kParameters</span><span class="p">,</span> <span class="s">&quot;Parameters are duplicated over different devices&quot;</span><span class="p">)</span>
    <span class="p">.</span><span class="n">AsDuplicable</span><span class="p">();</span>
<span class="n">AddInput</span><span class="p">(</span><span class="n">kPlaces</span><span class="p">,</span> <span class="s">&quot;Devices used for parallel processing&quot;</span><span class="p">);</span>
<span class="n">AddOutput</span><span class="p">(</span><span class="n">kOutputs</span><span class="p">,</span> <span class="s">&quot;Outputs needed to be merged from different devices&quot;</span><span class="p">).</span><span class="n">AsDuplicable</span><span class="p">();</span>
<span class="n">AddOutput</span><span class="p">(</span><span class="n">kParallelScopes</span><span class="p">,</span>
          <span class="s">&quot;Scopes for all local variables in forward pass. One scope for each device&quot;</span><span class="p">);</span>
<span class="n">AddAttr</span><span class="o">&lt;</span><span class="n">framework</span><span class="o">::</span><span class="n">BlockDesc</span> <span class="o">*&gt;</span><span class="p">(</span><span class="n">kParallelBlock</span><span class="p">,</span>
                                <span class="s">&quot;List of operaters to be executed in parallel&quot;</span><span class="p">);</span>
</pre></div>
</div>
<p>A vanilla implementation of parallel_do can be shown as the following (<code class="docutils literal"><span class="pre">|</span></code> means single thread and
<code class="docutils literal"><span class="pre">||||</span></code> means multiple threads)</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">In</span> <span class="n">the</span> <span class="n">forward</span> <span class="k">pass</span>
  <span class="o">|</span>      <span class="n">Split</span> <span class="nb">input</span> <span class="n">onto</span> <span class="n">different</span> <span class="n">devices</span>
220
  <span class="o">|</span>      <span class="n">Copy</span> <span class="n">parameter</span> <span class="n">onto</span> <span class="n">different</span> <span class="n">devices</span>
221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279
  <span class="o">||||</span>   <span class="n">Compute</span> <span class="n">forward</span> <span class="k">pass</span> <span class="ow">in</span> <span class="n">parallel</span>
  <span class="o">|</span>      <span class="n">Merge</span> <span class="n">output</span> <span class="kn">from</span> <span class="nn">different</span> <span class="n">devices</span>

<span class="n">In</span> <span class="n">the</span> <span class="n">backward</span> <span class="k">pass</span>
  <span class="o">|</span>      <span class="n">Split</span> <span class="n">output</span><span class="nd">@grad</span> <span class="n">onto</span> <span class="n">different</span> <span class="n">devices</span>
  <span class="o">||||</span>   <span class="n">Compute</span> <span class="n">backward</span> <span class="k">pass</span> <span class="ow">in</span> <span class="n">parallel</span>
  <span class="o">|</span>      <span class="n">accumulate</span> <span class="n">param</span><span class="nd">@grad</span> <span class="kn">from</span> <span class="nn">different</span> <span class="n">devices</span> <span class="n">to</span> <span class="n">the</span> <span class="n">first</span> <span class="n">device</span>
  <span class="o">|</span>      <span class="n">Merge</span> <span class="nb">input</span><span class="nd">@grad</span> <span class="kn">from</span> <span class="nn">different</span> <span class="n">devices</span>
  <span class="o">|</span>      <span class="n">Copy</span> <span class="n">param</span><span class="nd">@grad</span> <span class="n">to</span> <span class="n">the</span> <span class="n">place</span> <span class="n">of</span> <span class="n">parallel_do_op</span>
</pre></div>
</div>
<p>This implementation allows to write mixed device program like this</p>
<div class="highlight-python"><div class="highlight"><pre><span></span><span class="c1"># get embedding feature on CPU</span>
<span class="n">feature</span> <span class="o">=</span> <span class="n">some_cpu_only_op</span><span class="p">(</span><span class="n">data</span><span class="p">)</span>

<span class="n">gpu_places</span> <span class="o">=</span> <span class="n">get_place</span><span class="p">(</span><span class="n">use_gpu</span><span class="o">=</span><span class="bp">True</span><span class="p">)</span>
<span class="c1"># parallel processing on multiple GPUs</span>
<span class="n">pd</span> <span class="o">=</span> <span class="n">ParallelDo</span><span class="p">(</span><span class="n">gpu_places</span><span class="p">)</span>
<span class="k">with</span> <span class="n">pd</span><span class="o">.</span><span class="n">do</span><span class="p">():</span>
    <span class="n">read_input</span><span class="p">(</span><span class="n">feature</span><span class="p">)</span>
    <span class="n">prediction</span> <span class="o">=</span> <span class="n">my_net</span><span class="p">(</span><span class="n">feature</span><span class="p">)</span>
    <span class="n">write_output</span><span class="p">(</span><span class="n">prediction</span><span class="p">)</span>
<span class="n">prediction</span> <span class="o">=</span> <span class="n">pd</span><span class="p">()</span>
<span class="n">loss</span> <span class="o">=</span> <span class="n">cross_entropy</span><span class="p">(</span><span class="n">prediction</span><span class="p">,</span> <span class="n">label</span><span class="p">)</span>
</pre></div>
</div>
<p>And the programDesc are like the following</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="c1"># start_program will be run by executor(CPUPlace), all w1, w2 will be allocated on CPU</span>
<span class="n">start_program</span>
<span class="p">{</span>
  <span class="nb">vars</span><span class="p">:</span> <span class="n">w1</span><span class="p">,</span> <span class="n">w2</span>
  <span class="n">ops</span><span class="p">:</span> <span class="n">init</span><span class="p">(</span><span class="n">w1</span><span class="p">),</span> <span class="n">init</span><span class="p">(</span><span class="n">w2</span><span class="p">)</span>
<span class="p">}</span>

<span class="n">main_program</span>
<span class="p">{</span>
<span class="n">block0</span> <span class="p">{</span>
  <span class="nb">vars</span><span class="p">:</span> <span class="n">data</span><span class="p">,</span> <span class="n">places</span><span class="p">,</span> <span class="n">w1</span><span class="p">,</span> <span class="n">w2</span>
  <span class="n">ops</span><span class="p">:</span> <span class="n">data</span><span class="p">,</span> <span class="n">get_place</span><span class="p">,</span> <span class="n">parallel_do</span><span class="p">(</span><span class="n">block1</span><span class="p">),</span>
       <span class="n">parallel_do_grad</span><span class="p">(</span><span class="n">block2</span><span class="p">),</span>
       <span class="n">sgd</span><span class="p">(</span><span class="n">w2</span><span class="p">,</span> <span class="n">w2_grad</span><span class="p">),</span>
       <span class="n">sgd</span><span class="p">(</span><span class="n">w1</span><span class="p">,</span> <span class="n">w1_grad</span><span class="p">)</span>
<span class="p">}</span>
<span class="n">block1</span> <span class="p">{</span>
  <span class="n">parent_block</span><span class="p">:</span> <span class="mi">0</span>
  <span class="nb">vars</span><span class="p">:</span> <span class="n">data</span><span class="p">,</span> <span class="n">h1</span><span class="p">,</span> <span class="n">h2</span><span class="p">,</span> <span class="n">loss</span>
  <span class="n">ops</span><span class="p">:</span> <span class="n">fc</span><span class="p">,</span> <span class="n">fc</span><span class="p">,</span> <span class="n">softmax</span>
<span class="p">}</span>
<span class="n">block2</span> <span class="p">{</span>
  <span class="n">parent_block</span><span class="p">:</span> <span class="mi">1</span>
  <span class="nb">vars</span><span class="p">:</span> <span class="n">data_grad</span><span class="p">,</span> <span class="n">h1_grad</span><span class="p">,</span> <span class="n">h2_grad</span><span class="p">,</span> <span class="n">loss_gard</span><span class="p">,</span> <span class="n">w1_grad</span><span class="p">,</span> <span class="n">w2_grad</span>
  <span class="n">ops</span><span class="p">:</span> <span class="n">softmax_grad</span><span class="p">,</span>
       <span class="n">fc_grad</span>
       <span class="n">fc_grad</span>
<span class="p">}</span>
<span class="p">}</span>
</pre></div>
</div>
</div>
280 281
<div class="section" id="performance-imporvement">
<span id="performance-imporvement"></span><h2>Performance Imporvement<a class="headerlink" href="#performance-imporvement" title="永久链接至标题"></a></h2>
282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414
<p>There are serial places we can make this parallel_do faster.</p>
<div class="section" id="forward-split-input-onto-different-devices">
<span id="forward-split-input-onto-different-devices"></span><h3>forward: split input onto different devices<a class="headerlink" href="#forward-split-input-onto-different-devices" title="永久链接至标题"></a></h3>
<p>If the input of the parallel_do is independent from any prior opeartors, we can avoid this step by
prefetching the input onto different devices in a seperate background thread. And the python code
looks like this.</p>
<div class="highlight-python"><div class="highlight"><pre><span></span>pd = ParallelDo(gpu_places)
with pd.do():
    feature = get_data_from_prefetch_queue(gpu_places)
    prediction = my_net(feature)
    write_output(activation)
</pre></div>
</div>
</div>
<div class="section" id="forward-copy-parameter-to-onto-different-devices">
<span id="forward-copy-parameter-to-onto-different-devices"></span><h3>forward: Copy parameter to onto different devices<a class="headerlink" href="#forward-copy-parameter-to-onto-different-devices" title="永久链接至标题"></a></h3>
<p>We can avoid this step by making each device have a copy of the parameter. This requires:</p>
<ol class="simple">
<li><code class="docutils literal"><span class="pre">fluid.default_start_up_program()</span></code> to be run on all devices</li>
<li>In the backward, allreduce param&#64;grad at different devices, this requires<ol>
<li><code class="docutils literal"><span class="pre">backward.py</span></code> add <code class="docutils literal"><span class="pre">allreduce</span></code> operators at parallel_do_grad</li>
<li><code class="docutils literal"><span class="pre">allreduce</span></code> operators need to be called in async mode to achieve maximum throughput</li>
</ol>
</li>
<li>apply gradients related op(i.e. cliping, normalization, decay, sgd) on different devices in parallel</li>
</ol>
<p>By doing so, we also avoided &#8220;backward: accumulate param&#64;grad from different devices to the first device&#8221;.
And the ProgramDesc looks like the following</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="c1"># w1, w2 will be allocated on all GPUs</span>
<span class="n">start_program</span>
<span class="p">{</span>
<span class="n">block0</span> <span class="p">{</span>
  <span class="n">parallel_do</span><span class="p">(</span><span class="n">block1</span><span class="p">)</span>
<span class="p">}</span>
<span class="n">block1</span> <span class="p">{</span>
  <span class="n">parent_block</span><span class="p">:</span> <span class="mi">0</span>
  <span class="nb">vars</span><span class="p">:</span> <span class="n">w1</span><span class="p">,</span> <span class="n">w2</span>
  <span class="n">ops</span><span class="p">:</span> <span class="n">init</span><span class="p">(</span><span class="n">w1</span><span class="p">),</span> <span class="n">init</span><span class="p">(</span><span class="n">w2</span><span class="p">)</span>
<span class="p">}</span>
<span class="p">}</span>

<span class="n">main_program</span>
<span class="p">{</span>
<span class="n">block0</span> <span class="p">{</span>
  <span class="nb">vars</span><span class="p">:</span> <span class="n">data</span><span class="p">,</span> <span class="n">places</span><span class="p">,</span> <span class="n">w1</span><span class="p">,</span> <span class="n">w2</span>
  <span class="n">ops</span><span class="p">:</span> <span class="n">data</span><span class="p">,</span> <span class="n">get_place</span><span class="p">,</span> <span class="n">parallel_do</span><span class="p">(</span><span class="n">block1</span><span class="p">),</span>
       <span class="n">parallel_do_grad</span><span class="p">(</span><span class="n">block2</span><span class="p">),</span>      <span class="c1"># append_backward</span>
       <span class="n">parallel_do</span><span class="p">(</span><span class="n">block3</span><span class="p">)</span>            <span class="c1"># append_optimization</span>
       
<span class="p">}</span>
<span class="n">block1</span> <span class="p">{</span>
  <span class="n">parent_block</span><span class="p">:</span> <span class="mi">0</span>
  <span class="nb">vars</span><span class="p">:</span> <span class="n">data</span><span class="p">,</span> <span class="n">h1</span><span class="p">,</span> <span class="n">h2</span><span class="p">,</span> <span class="n">loss</span>
  <span class="n">ops</span><span class="p">:</span> <span class="n">fc</span><span class="p">,</span> <span class="n">fc</span><span class="p">,</span> <span class="n">softmax</span>
<span class="p">}</span>
<span class="n">block2</span> <span class="p">{</span>
  <span class="n">parent_block</span><span class="p">:</span> <span class="mi">1</span>
  <span class="nb">vars</span><span class="p">:</span> <span class="n">data_grad</span><span class="p">,</span> <span class="n">h1_grad</span><span class="p">,</span> <span class="n">h2_grad</span><span class="p">,</span> <span class="n">loss_gard</span><span class="p">,</span> <span class="n">w1_grad</span><span class="p">,</span> <span class="n">w2_grad</span>
  <span class="n">ops</span><span class="p">:</span> <span class="n">softmax_grad</span><span class="p">,</span>
       <span class="n">fc_grad</span><span class="p">,</span> <span class="n">allreduce</span><span class="p">(</span><span class="n">places</span><span class="p">,</span> <span class="n">scopes</span><span class="p">,</span> <span class="n">w1_grad</span><span class="p">),</span>
       <span class="n">fc_grad</span><span class="p">,</span> <span class="n">allreduce</span><span class="p">(</span><span class="n">places</span><span class="p">,</span> <span class="n">scopes</span><span class="p">,</span> <span class="n">w2_grad</span><span class="p">)</span>
<span class="p">}</span>
<span class="n">block3</span> <span class="p">{</span>
  <span class="n">parent_block</span><span class="p">:</span> <span class="mi">0</span>
  <span class="nb">vars</span><span class="p">:</span> <span class="n">lr</span>
  <span class="n">ops</span><span class="p">:</span> <span class="n">sgd</span><span class="p">(</span><span class="n">w2</span><span class="p">,</span> <span class="n">w2_grad</span><span class="p">),</span>
       <span class="n">sgd</span><span class="p">(</span><span class="n">w1</span><span class="p">,</span> <span class="n">w1_grad</span><span class="p">)</span>
<span class="p">}</span>
<span class="p">}</span>
</pre></div>
</div>
</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,
            SOURCELINK_SUFFIX: ".txt",
        };
    </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="../_static/translations.js"></script>
      <script type="text/javascript" src="https://cdn.bootcss.com/mathjax/2.7.0/MathJax.js"></script>
       
  

  
  
    <script type="text/javascript" src="../_static/js/theme.js"></script>
  
  
  <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
  <script src="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/js/perfect-scrollbar.jquery.min.js"></script>
  <script src="../_static/js/paddle_doc_init.js"></script> 

</body>
</html>