use_case_en.html 31.2 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


<!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>Use Case &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"/>
        <link rel="up" title="Set Command-line Parameters" href="index_en.html"/>
        <link rel="next" title="Argument Outline" href="arguments_en.html"/>
        <link rel="prev" title="Set Command-line Parameters" href="index_en.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>Folk 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>Home</a></li>
          <li><a>Get Started</a></li>
          <li class="active"><a>Documentation</a></li>
          <li><a>About Us</a></li>
        </ul>
      </div>
      <div class="doc-module">
        
        <ul class="current">
<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_en.html">GET STARTED</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../../tutorials/index_en.html">TUTORIALS</a></li>
<li class="toctree-l1 current"><a class="reference internal" href="../../index_en.html">HOW TO</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../../api/index_en.html">API</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../../about/index_en.html">ABOUT</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 class="current">
<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/build_and_install/index_en.html">Install and Build</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/docker_install_en.html">PaddlePaddle in Docker Containers</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/ubuntu_install_en.html">Debian Package installation guide</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/build_from_source_en.html">Installing from Sources</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../../../getstarted/basic_usage/index_en.html">Simple Linear Regression</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../../tutorials/index_en.html">TUTORIALS</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../../../tutorials/quick_start/index_en.html">Quick Start</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../../tutorials/rec/ml_regression_en.html">MovieLens Regression</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../../tutorials/image_classification/index_en.html">Image Classification</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../../tutorials/sentiment_analysis/index_en.html">Sentiment Analysis</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../../tutorials/semantic_role_labeling/index_en.html">Semantic Role Labeling</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../../tutorials/text_generation/index_en.html">Text Generation</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../../tutorials/gan/index_en.html">Image Auto-Generation</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../../tutorials/imagenet_model/resnet_model_en.html">ImageNet: ResNet</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../../tutorials/embedding_model/index_en.html">Embedding: Chinese Word</a></li>
</ul>
</li>
<li class="toctree-l1 current"><a class="reference internal" href="../../index_en.html">HOW TO</a><ul class="current">
<li class="toctree-l2 current"><a class="reference internal" href="index_en.html">Set Command-line Parameters</a><ul class="current">
<li class="toctree-l3 current"><a class="current reference internal" href="#">Use Case</a></li>
<li class="toctree-l3"><a class="reference internal" href="arguments_en.html">Argument Outline</a></li>
<li class="toctree-l3"><a class="reference internal" href="detail_introduction_en.html">Detail Description</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../cluster/cluster_train_en.html">Run Distributed Training</a></li>
<li class="toctree-l2"><a class="reference internal" href="../k8s/k8s_en.html">Paddle On Kubernetes</a></li>
<li class="toctree-l2"><a class="reference internal" href="../k8s/k8s_aws_en.html">Distributed PaddlePaddle Training on AWS with Kubernetes</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../dev/new_layer_en.html">Write New Layers</a></li>
<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="../../deep_model/rnn/index_en.html">RNN Models</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../../deep_model/rnn/rnn_config_en.html">RNN Configuration</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../../optimization/gpu_profiling_en.html">Tune GPU Performance</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../../api/index_en.html">API</a><ul>
158 159 160 161 162 163 164 165 166 167 168
<li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">Model Configuration</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/attr.html">Parameter Attribute</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../../../api/v2/data.html">Datasets</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../../api/v2/run_logic.html">Training and Inference</a></li>
169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../../about/index_en.html">ABOUT</a></li>
</ul>

        
    </nav>
    
    <nav class="local-toc"><ul>
<li><a class="reference internal" href="#">Use Case</a><ul>
<li><a class="reference internal" href="#local-training">Local Training</a><ul>
<li><a class="reference internal" href="#pass-command-argument-to-network-config">1) Pass Command Argument to Network config</a></li>
<li><a class="reference internal" href="#use-model-to-initialize-network">2) Use Model to Initialize Network</a></li>
</ul>
</li>
<li><a class="reference internal" href="#local-testing">Local Testing</a></li>
<li><a class="reference internal" href="#sparse-training">Sparse Training</a><ul>
<li><a class="reference internal" href="#local-training">1) Local training</a></li>
<li><a class="reference internal" href="#cluster-training">2) cluster training</a></li>
</ul>
</li>
<li><a class="reference internal" href="#parallel-nn">parallel_nn</a><ul>
<li><a class="reference internal" href="#case-1-mixed-use-of-gpu-and-cpu">case 1: Mixed Use of GPU and CPU</a></li>
<li><a class="reference internal" href="#case-2-specify-layers-in-different-devices">Case 2: Specify Layers in Different Devices</a></li>
</ul>
</li>
</ul>
</li>
</ul>
</nav>
    
    <section class="doc-content-wrap">

      

 







<div role="navigation" aria-label="breadcrumbs navigation">
  <ul class="wy-breadcrumbs">
      
        <li><a href="../../index_en.html">HOW TO</a> > </li>
      
        <li><a href="index_en.html">Set Command-line Parameters</a> > </li>
      
    <li>Use Case</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="use-case">
<span id="use-case"></span><h1>Use Case<a class="headerlink" href="#use-case" title="Permalink to this headline"></a></h1>
<div class="section" id="local-training">
<span id="local-training"></span><h2>Local Training<a class="headerlink" href="#local-training" title="Permalink to this headline"></a></h2>
<p>These command line arguments are commonly used by local training experiments, such as image classification, natural language processing, et al.</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">paddle</span> <span class="n">train</span> \
  <span class="o">--</span><span class="n">use_gpu</span><span class="o">=</span><span class="mi">1</span><span class="o">/</span><span class="mi">0</span> \                        <span class="c1">#1:GPU,0:CPU(default:true)</span>
  <span class="o">--</span><span class="n">config</span><span class="o">=</span><span class="n">network_config</span> \
  <span class="o">--</span><span class="n">save_dir</span><span class="o">=</span><span class="n">output</span> \
  <span class="o">--</span><span class="n">trainer_count</span><span class="o">=</span><span class="n">COUNT</span> \                <span class="c1">#(default:1)</span>
  <span class="o">--</span><span class="n">test_period</span><span class="o">=</span><span class="n">M</span> \                      <span class="c1">#(default:0) </span>
  <span class="o">--</span><span class="n">num_passes</span><span class="o">=</span><span class="n">N</span> \                       <span class="c1">#(defalut:100)</span>
  <span class="o">--</span><span class="n">log_period</span><span class="o">=</span><span class="n">K</span> \                       <span class="c1">#(default:100)</span>
  <span class="o">--</span><span class="n">dot_period</span><span class="o">=</span><span class="mi">1000</span> \                    <span class="c1">#(default:1)</span>
  <span class="c1">#[--show_parameter_stats_period=100] \ #(default:0)</span>
  <span class="c1">#[--saving_period_by_batches=200] \    #(default:0)</span>
</pre></div>
</div>
<p><code class="docutils literal"><span class="pre">show_parameter_stats_period</span></code> and <code class="docutils literal"><span class="pre">saving_period_by_batches</span></code> are optional according to your task.</p>
<div class="section" id="pass-command-argument-to-network-config">
<span id="pass-command-argument-to-network-config"></span><h3>1) Pass Command Argument to Network config<a class="headerlink" href="#pass-command-argument-to-network-config" title="Permalink to this headline"></a></h3>
<p><code class="docutils literal"><span class="pre">config_args</span></code> is a useful parameter to pass arguments to network config.</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="o">--</span><span class="n">config_args</span><span class="o">=</span><span class="n">generating</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span><span class="n">beam_size</span><span class="o">=</span><span class="mi">5</span><span class="p">,</span><span class="n">layer_num</span><span class="o">=</span><span class="mi">10</span> \
</pre></div>
</div>
<p>And <code class="docutils literal"><span class="pre">get_config_arg</span></code> can be used to parse these arguments in network config as follows:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">generating</span> <span class="o">=</span> <span class="n">get_config_arg</span><span class="p">(</span><span class="s1">&#39;generating&#39;</span><span class="p">,</span> <span class="nb">bool</span><span class="p">,</span> <span class="kc">False</span><span class="p">)</span>
<span class="n">beam_size</span> <span class="o">=</span> <span class="n">get_config_arg</span><span class="p">(</span><span class="s1">&#39;beam_size&#39;</span><span class="p">,</span> <span class="nb">int</span><span class="p">,</span> <span class="mi">3</span><span class="p">)</span>
<span class="n">layer_num</span> <span class="o">=</span> <span class="n">get_config_arg</span><span class="p">(</span><span class="s1">&#39;layer_num&#39;</span><span class="p">,</span> <span class="nb">int</span><span class="p">,</span> <span class="mi">8</span><span class="p">)</span>
</pre></div>
</div>
<p><code class="docutils literal"><span class="pre">get_config_arg</span></code>:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">get_config_arg</span><span class="p">(</span><span class="n">name</span><span class="p">,</span> <span class="nb">type</span><span class="p">,</span> <span class="n">default_value</span><span class="p">)</span>
</pre></div>
</div>
<ul class="simple">
<li>name: the name specified in the <code class="docutils literal"><span class="pre">--config_args</span></code></li>
<li>type: value type, bool, int, str, float etc.</li>
<li>default_value: default value if not set.</li>
</ul>
</div>
<div class="section" id="use-model-to-initialize-network">
<span id="use-model-to-initialize-network"></span><h3>2) Use Model to Initialize Network<a class="headerlink" href="#use-model-to-initialize-network" title="Permalink to this headline"></a></h3>
<p>add argument:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="o">--</span><span class="n">init_model_path</span><span class="o">=</span><span class="n">model_path</span>
<span class="o">--</span><span class="n">load_missing_parameter_strategy</span><span class="o">=</span><span class="n">rand</span>
</pre></div>
</div>
</div>
</div>
<div class="section" id="local-testing">
<span id="local-testing"></span><h2>Local Testing<a class="headerlink" href="#local-testing" title="Permalink to this headline"></a></h2>
<p>Method 1:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">paddle</span> <span class="n">train</span> <span class="o">--</span><span class="n">job</span><span class="o">=</span><span class="n">test</span> \
             <span class="o">--</span><span class="n">use_gpu</span><span class="o">=</span><span class="mi">1</span><span class="o">/</span><span class="mi">0</span> \ 
             <span class="o">--</span><span class="n">config</span><span class="o">=</span><span class="n">network_config</span> \
             <span class="o">--</span><span class="n">trainer_count</span><span class="o">=</span><span class="n">COUNT</span> \ 
             <span class="o">--</span><span class="n">init_model_path</span><span class="o">=</span><span class="n">model_path</span> \
</pre></div>
</div>
<ul class="simple">
<li>use init_model_path to specify test model.</li>
<li>only can test one model.</li>
</ul>
<p>Method 2:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">paddle</span> <span class="n">train</span> <span class="o">--</span><span class="n">job</span><span class="o">=</span><span class="n">test</span> \
             <span class="o">--</span><span class="n">use_gpu</span><span class="o">=</span><span class="mi">1</span><span class="o">/</span><span class="mi">0</span> \ 
             <span class="o">--</span><span class="n">config</span><span class="o">=</span><span class="n">network_config</span> \
             <span class="o">--</span><span class="n">trainer_count</span><span class="o">=</span><span class="n">COUNT</span> \ 
             <span class="o">--</span><span class="n">model_list</span><span class="o">=</span><span class="n">model</span><span class="o">.</span><span class="n">list</span> \
</pre></div>
</div>
<ul class="simple">
<li>use model_list to specify test models</li>
<li>can test several models, where model.list likes:</li>
</ul>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="o">./</span><span class="n">alexnet_pass1</span>
<span class="o">./</span><span class="n">alexnet_pass2</span>
</pre></div>
</div>
<p>Method 3:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">paddle</span> <span class="n">train</span> <span class="o">--</span><span class="n">job</span><span class="o">=</span><span class="n">test</span> \
             <span class="o">--</span><span class="n">use_gpu</span><span class="o">=</span><span class="mi">1</span><span class="o">/</span><span class="mi">0</span> \
             <span class="o">--</span><span class="n">config</span><span class="o">=</span><span class="n">network_config</span> \
             <span class="o">--</span><span class="n">trainer_count</span><span class="o">=</span><span class="n">COUNT</span> \
             <span class="o">--</span><span class="n">save_dir</span><span class="o">=</span><span class="n">model</span> \
             <span class="o">--</span><span class="n">test_pass</span><span class="o">=</span><span class="n">M</span> \
             <span class="o">--</span><span class="n">num_passes</span><span class="o">=</span><span class="n">N</span> \
</pre></div>
</div>
<p>This way must use model path saved by Paddle like this: <code class="docutils literal"><span class="pre">model/pass-%5d</span></code>. Testing model is from M-th pass to (N-1)-th pass. For example: M=12 and N=14 will test <code class="docutils literal"><span class="pre">model/pass-00012</span></code> and <code class="docutils literal"><span class="pre">model/pass-00013</span></code>.</p>
</div>
<div class="section" id="sparse-training">
<span id="sparse-training"></span><h2>Sparse Training<a class="headerlink" href="#sparse-training" title="Permalink to this headline"></a></h2>
<p>Sparse training is usually used to accelerate calculation when input is sparse data with highly dimension. For example, dictionary dimension of input data is 1 million, but one sample just have several words. In paddle, sparse matrix multiplication is used in forward propagation and sparse updating is perfomed on weight updating after backward propagation.</p>
<div class="section" id="local-training">
<span id="id1"></span><h3>1) Local training<a class="headerlink" href="#local-training" title="Permalink to this headline"></a></h3>
<p>You need to set <strong>sparse_update=True</strong> in network config.  Check the network config documentation for more details.</p>
</div>
<div class="section" id="cluster-training">
<span id="cluster-training"></span><h3>2) cluster training<a class="headerlink" href="#cluster-training" title="Permalink to this headline"></a></h3>
<p>Add the following argument for cluster training of a sparse model. At the same time you need to set <strong>sparse_remote_update=True</strong> in network config. Check the network config documentation for more details.</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="o">--</span><span class="n">ports_num_for_sparse</span><span class="o">=</span><span class="mi">1</span>    <span class="c1">#(default: 0)</span>
</pre></div>
</div>
</div>
</div>
<div class="section" id="parallel-nn">
<span id="parallel-nn"></span><h2>parallel_nn<a class="headerlink" href="#parallel-nn" title="Permalink to this headline"></a></h2>
<p><code class="docutils literal"><span class="pre">parallel_nn</span></code> can be set to mixed use of GPUs and CPUs to compute layers. That is to say, you can deploy network to use a GPU to compute some layers and use a CPU to compute other layers. The other way is to split layers into different GPUs, which can <strong>reduce GPU memory</strong> or <strong>use parallel computation to accelerate some layers</strong>.</p>
<p>If you want to use these characteristics, you need to specify device ID in network config (denote it as deviceId) and add command line argument:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="o">--</span><span class="n">parallel_nn</span><span class="o">=</span><span class="n">true</span>
</pre></div>
</div>
<div class="section" id="case-1-mixed-use-of-gpu-and-cpu">
<span id="case-1-mixed-use-of-gpu-and-cpu"></span><h3>case 1: Mixed Use of GPU and CPU<a class="headerlink" href="#case-1-mixed-use-of-gpu-and-cpu" title="Permalink to this headline"></a></h3>
<p>Consider the following example:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="c1">#command line:</span>
<span class="n">paddle</span> <span class="n">train</span> <span class="o">--</span><span class="n">use_gpu</span><span class="o">=</span><span class="n">true</span> <span class="o">--</span><span class="n">parallel_nn</span><span class="o">=</span><span class="n">true</span> <span class="n">trainer_count</span><span class="o">=</span><span class="n">COUNT</span>

<span class="n">default_device</span><span class="p">(</span><span class="mi">0</span><span class="p">)</span>

<span class="n">fc1</span><span class="o">=</span><span class="n">fc_layer</span><span class="p">(</span><span class="o">...</span><span class="p">)</span>
<span class="n">fc2</span><span class="o">=</span><span class="n">fc_layer</span><span class="p">(</span><span class="o">...</span><span class="p">)</span>
<span class="n">fc3</span><span class="o">=</span><span class="n">fc_layer</span><span class="p">(</span><span class="o">...</span><span class="p">,</span><span class="n">layer_attr</span><span class="o">=</span><span class="n">ExtraAttr</span><span class="p">(</span><span class="n">device</span><span class="o">=-</span><span class="mi">1</span><span class="p">))</span>
</pre></div>
</div>
<ul class="simple">
<li>default_device(0): set default device ID to 0. This means that except the layers with device=-1, all layers will use a GPU, and the specific GPU used for each layer depends on trainer_count and gpu_id (0 by default). Here, layer fc1 and fc2 are computed on the GPU.</li>
<li>device=-1: use the CPU for layer fc3.</li>
<li>trainer_count:<ul>
<li>trainer_count=1: if gpu_id is not set, then use the first GPU to compute layers fc1 and fc2. Otherwise use the GPU with gpu_id.</li>
<li>trainer_count&gt;1: use trainer_count GPUs to compute one layer using data parallelism. For example, trainer_count=2 means that GPUs 0 and 1 will use data parallelism to compute layer fc1 and fc2.</li>
</ul>
</li>
</ul>
</div>
<div class="section" id="case-2-specify-layers-in-different-devices">
<span id="case-2-specify-layers-in-different-devices"></span><h3>Case 2: Specify Layers in Different Devices<a class="headerlink" href="#case-2-specify-layers-in-different-devices" title="Permalink to this headline"></a></h3>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="c1">#command line:</span>
<span class="n">paddle</span> <span class="n">train</span> <span class="o">--</span><span class="n">use_gpu</span><span class="o">=</span><span class="n">true</span> <span class="o">--</span><span class="n">parallel_nn</span><span class="o">=</span><span class="n">true</span> <span class="o">--</span><span class="n">trainer_count</span><span class="o">=</span><span class="n">COUNT</span>

<span class="c1">#network:</span>
<span class="n">fc2</span><span class="o">=</span><span class="n">fc_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">l1</span><span class="p">,</span> <span class="n">layer_attr</span><span class="o">=</span><span class="n">ExtraAttr</span><span class="p">(</span><span class="n">device</span><span class="o">=</span><span class="mi">0</span><span class="p">),</span> <span class="o">...</span><span class="p">)</span>
<span class="n">fc3</span><span class="o">=</span><span class="n">fc_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">l1</span><span class="p">,</span> <span class="n">layer_attr</span><span class="o">=</span><span class="n">ExtraAttr</span><span class="p">(</span><span class="n">device</span><span class="o">=</span><span class="mi">1</span><span class="p">),</span> <span class="o">...</span><span class="p">)</span>
<span class="n">fc4</span><span class="o">=</span><span class="n">fc_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">fc2</span><span class="p">,</span> <span class="n">layer_attr</span><span class="o">=</span><span class="n">ExtraAttr</span><span class="p">(</span><span class="n">device</span><span class="o">=-</span><span class="mi">1</span><span class="p">),</span> <span class="o">...</span><span class="p">)</span>
</pre></div>
</div>
<p>In this case, we assume that there are 4 GPUs in one machine.</p>
<ul class="simple">
<li>trainer_count=1:<ul>
<li>Use GPU 0 to compute layer fc2.</li>
<li>Use GPU 1 to compute layer fc3.</li>
<li>Use CPU to compute layer fc4.</li>
</ul>
</li>
<li>trainer_count=2:<ul>
<li>Use GPU 0 and 1 to compute layer fc2.</li>
<li>Use GPU 2 and 3 to compute layer fc3.</li>
<li>Use CPU to compute fc4 in two threads.</li>
</ul>
</li>
<li>trainer_count=4:<ul>
<li>It will fail (note, we have assumed that there are 4 GPUs in machine), because argument <code class="docutils literal"><span class="pre">allow_only_one_model_on_one_gpu</span></code> is true by default.</li>
</ul>
</li>
</ul>
<p><strong>Allocation of device ID when <code class="docutils literal"><span class="pre">device!=-1</span></code></strong>:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="p">(</span><span class="n">deviceId</span> <span class="o">+</span> <span class="n">gpu_id</span> <span class="o">+</span> <span class="n">threadId</span> <span class="o">*</span> <span class="n">numLogicalDevices_</span><span class="p">)</span> <span class="o">%</span> <span class="n">numDevices_</span>

<span class="n">deviceId</span><span class="p">:</span>             <span class="n">specified</span> <span class="ow">in</span> <span class="n">layer</span><span class="o">.</span>
<span class="n">gpu_id</span><span class="p">:</span>               <span class="mi">0</span> <span class="n">by</span> <span class="n">default</span><span class="o">.</span>
<span class="n">threadId</span><span class="p">:</span>             <span class="n">thread</span> <span class="n">ID</span><span class="p">,</span> <span class="nb">range</span><span class="p">:</span> <span class="mi">0</span><span class="p">,</span><span class="mi">1</span><span class="p">,</span><span class="o">...</span><span class="p">,</span> <span class="n">trainer_count</span><span class="o">-</span><span class="mi">1</span>
<span class="n">numDevices_</span><span class="p">:</span>          <span class="n">device</span> <span class="p">(</span><span class="n">GPU</span><span class="p">)</span> <span class="n">count</span> <span class="ow">in</span> <span class="n">machine</span><span class="o">.</span>
<span class="n">numLogicalDevices_</span><span class="p">:</span>   <span class="nb">min</span><span class="p">(</span><span class="nb">max</span><span class="p">(</span><span class="n">deviceId</span> <span class="o">+</span> <span class="mi">1</span><span class="p">),</span> <span class="n">numDevices_</span><span class="p">)</span>
</pre></div>
</div>
</div>
</div>
</div>


           </div>
          </div>
          <footer>
  
    <div class="rst-footer-buttons" role="navigation" aria-label="footer navigation">
      
        <a href="arguments_en.html" class="btn btn-neutral float-right" title="Argument Outline" accesskey="n">Next <span class="fa fa-arrow-circle-right"></span></a>
      
      
        <a href="index_en.html" class="btn btn-neutral" title="Set Command-line Parameters" accesskey="p"><span class="fa fa-arrow-circle-left"></span> Previous</a>
      
    </div>
  

  <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://cdn.mathjax.org/mathjax/latest/MathJax.js?config=TeX-AMS-MML_HTMLorMML"></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>