Skip to content

P
PaddleSlim

PaddlePaddle / PaddleSlim
1 年多 前同步成功

通知 51
Star 1434
Fork 344
P
PaddleSlim
切换分支/标签
查找文件 普通视图历史永久链接Permalink
index.html 27.4 KB
Newer Older
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
1 2 3 4 5 6 7 8 9 
<!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 http-equiv="X-UA-Compatible" content="IE=edge">
  <meta name="viewport" content="width=device-width, initial-scale=1.0">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
10 
  <link rel="shortcut icon" href="../img/favicon.ico">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
11 
  <title>搜索空间 - PaddleSlim Docs</title>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
12 13 
  <link href='https://fonts.googleapis.com/css?family=Lato:400,700|Roboto+Slab:400,700|Inconsolata:400,700' rel='stylesheet' type='text/css'>
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
14 15 
  <link rel="stylesheet" href="../css/theme.css" type="text/css" />
  <link rel="stylesheet" href="../css/theme_extra.css" type="text/css" />
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
16 17 18 19 
  <link rel="stylesheet" href="//cdnjs.cloudflare.com/ajax/libs/highlight.js/9.12.0/styles/github.min.css">
  
  <script>
    // Current page data
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
20 
    var mkdocs_page_name = "\u641c\u7d22\u7a7a\u95f4";
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
21 
    var mkdocs_page_input_path = "search_space.md";
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
22 23 24 
    var mkdocs_page_url = null;
  </script>
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
25 26 
  <script src="../js/jquery-2.1.1.min.js" defer></script>
  <script src="../js/modernizr-2.8.3.min.js" defer></script>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
27 28 29 30 31 32 33 34 35 36 37 38 
  <script src="//cdnjs.cloudflare.com/ajax/libs/highlight.js/9.12.0/highlight.min.js"></script>
  <script>hljs.initHighlightingOnLoad();</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 stickynav">
      <div class="wy-side-nav-search">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
39 
        <a href=".." class="icon icon-home"> PaddleSlim Docs</a>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
40 
        <div role="search">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
41 
  <form id ="rtd-search-form" class="wy-form" action="../search.html" method="get">
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
42 43 44 45 46 47 48 49 50 51 52 
    <input type="text" name="q" placeholder="Search docs" title="Type search term here" />
  </form>
</div>
      </div>

      <div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="main navigation">
	<ul class="current">
	  
          
            <li class="toctree-l1">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
53 
    <a class="" href="..">Home</a>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
54 55 56 57 
	    </li>
          
            <li class="toctree-l1">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
58 
    <a class="" href="../model_zoo/">模型库</a>
B
Deployed 482d368 with MkDocs version: 1.0.4  
baiyfbupt 已提交
59 60 61 62 
	    </li>
          
            <li class="toctree-l1">
B
Deployed 55210fe with MkDocs version: 1.0.4  
baiyfbupt 已提交
63 64 65 66 
    <span class="caption-text">教程</span>
    <ul class="subnav">
                <li class="">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
67 
    <a class="" href="../tutorials/quant_post_demo/">离线量化</a>
B
Deployed 55210fe with MkDocs version: 1.0.4  
baiyfbupt 已提交
68 69 70 
                </li>
                <li class="">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
71 
    <a class="" href="../tutorials/quant_aware_demo/">量化训练</a>
B
Deployed 55210fe with MkDocs version: 1.0.4  
baiyfbupt 已提交
72 73 74 
                </li>
                <li class="">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
75 
    <a class="" href="../tutorials/quant_embedding_demo/">Embedding量化</a>
B
Deployed 55210fe with MkDocs version: 1.0.4  
baiyfbupt 已提交
76 77 78 
                </li>
                <li class="">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
79 
    <a class="" href="../tutorials/nas_demo/">SA搜索</a>
B
Deployed 55210fe with MkDocs version: 1.0.4  
baiyfbupt 已提交
80 81 82 
                </li>
                <li class="">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
83 
    <a class="" href="../tutorials/distillation_demo/">知识蒸馏</a>
B
Deployed 55210fe with MkDocs version: 1.0.4  
baiyfbupt 已提交
84 85 86 87 88 89 
                </li>
    </ul>
	    </li>
          
            <li class="toctree-l1">
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
90 91 92 93 
    <span class="caption-text">API</span>
    <ul class="subnav">
                <li class="">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
94 
    <a class="" href="../api/quantization_api/">量化</a>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
95 96 97 
                </li>
                <li class="">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
98 
    <a class="" href="../api/prune_api/">剪枝与敏感度</a>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
99 100 101 
                </li>
                <li class="">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
102 
    <a class="" href="../api/analysis_api/">模型分析</a>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
103 104 105 
                </li>
                <li class="">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
106 
    <a class="" href="../api/single_distiller_api/">知识蒸馏</a>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
107 108 109 
                </li>
                <li class="">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
110 
    <a class="" href="../api/nas_api/">SA搜索</a>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
111 
                </li>
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
112 
                <li class=" current">
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
113
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 
    <a class="current" href="./">搜索空间</a>
    <ul class="subnav">
            
    <li class="toctree-l3"><a href="#_1">搜索空间简介</a></li>
    

    <li class="toctree-l3"><a href="#paddleslimnas">paddleslim.nas 提供的搜索空间</a></li>
    
        <ul>
        
            <li><a class="toctree-l4" href="#_2">根据初始模型结构构造搜索空间</a></li>
        
            <li><a class="toctree-l4" href="#block">根据相应模型的block构造搜索空间</a></li>
        
        </ul>
    

    <li class="toctree-l3"><a href="#_3">搜索空间示例</a></li>
    

    <li class="toctree-l3"><a href="#search-space">自定义搜索空间(search space)</a></li>
    

    </ul>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
138 
                </li>
B
Deployed 0f9c3ce with MkDocs version: 1.0.4  
baiyfbupt 已提交
139 140 
                <li class="">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
141 
    <a class="" href="../table_latency/">硬件延时评估表</a>
B
Deployed 0f9c3ce with MkDocs version: 1.0.4  
baiyfbupt 已提交
142 
                </li>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
143 144 145 
    </ul>
	    </li>
B
Deployed 0f9c3ce with MkDocs version: 1.0.4  
baiyfbupt 已提交
146 147 
            <li class="toctree-l1">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
148 
    <a class="" href="../algo/algo/">算法原理</a>
B
Deployed 0f9c3ce with MkDocs version: 1.0.4  
baiyfbupt 已提交
149 150 
	    </li>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
151 152 153 154 155 156 157 158 159 160 
        </ul>
      </div>
      &nbsp;
    </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>
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
161 
        <a href="..">PaddleSlim Docs</a>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
162 163 164 165 166 167 168 
      </nav>

      
      <div class="wy-nav-content">
        <div class="rst-content">
          <div role="navigation" aria-label="breadcrumbs navigation">
  <ul class="wy-breadcrumbs">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
169 
    <li><a href="..">Docs</a> &raquo;</li>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
170 171
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
172 173 174 175 
        
          <li>API &raquo;</li>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
176
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
177 
    <li>搜索空间</li>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
178 179 
    <li class="wy-breadcrumbs-aside">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
180 
        <a href="https://github.com/PaddlePaddle/PaddleSlim/blob/develop/docs/docs/search_space.md"
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
181 182 183 184 185 186 187 188 189 
          class="icon icon-github"> Edit on GitHub</a>
      
    </li>
  </ul>
  <hr/>
</div>
          <div role="main">
            <div class="section">
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
190 191 192 193 194 195 
                <h2 id="_1">搜索空间简介<a class="headerlink" href="#_1" title="Permanent link">#</a></h2>
<dl>
<dd>搜索空间是神经网络搜索中的一个概念。搜索空间是一系列模型结构的汇集, SANAS主要是利用模拟退火的思想在搜索空间中搜索到一个比较小的模型结构或者一个精度比较高的模型结构。</dd>
</dl>
<h2 id="paddleslimnas">paddleslim.nas 提供的搜索空间<a class="headerlink" href="#paddleslimnas" title="Permanent link">#</a></h2>
<h5 id="_2">根据初始模型结构构造搜索空间<a class="headerlink" href="#_2" title="Permanent link">#</a></h5>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
196 
<ol>
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
197 198 199 200 201 202 203 204 205 206 207 208 
<li>
<p>MobileNetV2Space<br>
&emsp; MobileNetV2的网络结构可以参考：<a href="https://github.com/PaddlePaddle/models/blob/develop/PaddleCV/image_classification/models/mobilenet_v2.py#L29">代码</a>，<a href="https://arxiv.org/abs/1801.04381">论文</a></p>
</li>
<li>
<p>MobileNetV1Space<br>
&emsp; MobilNetV1的网络结构可以参考：<a href="https://github.com/PaddlePaddle/models/blob/develop/PaddleCV/image_classification/models/mobilenet_v1.py#L29">代码</a>，<a href="https://arxiv.org/abs/1704.04861">论文</a></p>
</li>
<li>
<p>ResNetSpace<br>
&emsp; ResNetSpace的网络结构可以参考：<a href="https://github.com/PaddlePaddle/models/blob/develop/PaddleCV/image_classification/models/resnet.py#L30">代码</a>，<a href="https://arxiv.org/pdf/1512.03385.pdf">论文</a></p>
</li>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
209 
</ol>
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
210 
<h5 id="block">根据相应模型的block构造搜索空间<a class="headerlink" href="#block" title="Permanent link">#</a></h5>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
211 
<ol>
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 
<li>
<p>MobileNetV1BlockSpace<br>
&emsp; MobileNetV1Block的结构可以参考：<a href="https://github.com/PaddlePaddle/models/blob/develop/PaddleCV/image_classification/models/mobilenet_v1.py#L173">代码</a></p>
</li>
<li>
<p>MobileNetV2BlockSpace<br>
&emsp; MobileNetV2Block的结构可以参考：<a href="https://github.com/PaddlePaddle/models/blob/develop/PaddleCV/image_classification/models/mobilenet_v2.py#L174">代码</a></p>
</li>
<li>
<p>ResNetBlockSpace<br>
&emsp; ResNetBlock的结构可以参考：<a href="https://github.com/PaddlePaddle/models/blob/develop/PaddleCV/image_classification/models/resnet.py#L148">代码</a></p>
</li>
<li>
<p>InceptionABlockSpace<br>
&emsp; InceptionABlock的结构可以参考：<a href="https://github.com/PaddlePaddle/models/blob/develop/PaddleCV/image_classification/models/inception_v4.py#L140">代码</a></p>
</li>
<li>
<p>InceptionCBlockSpace<br>
&emsp; InceptionCBlock结构可以参考：<a href="https://github.com/PaddlePaddle/models/blob/develop/PaddleCV/image_classification/models/inception_v4.py#L291">代码</a></p>
</li>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
232 
</ol>
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
233 
<h2 id="_3">搜索空间示例<a class="headerlink" href="#_3" title="Permanent link">#</a></h2>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
234 
<ol>
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
235 236 237 
<li>使用paddleslim中提供用初始的模型结构来构造搜索空间的话，仅需要指定搜索空间名字即可。例如：如果使用原本的MobileNetV2的搜索空间进行搜索的话，传入SANAS中的config直接指定为[('MobileNetV2Space')]。</li>
<li>使用paddleslim中提供的block搜索空间构造搜索空间：<br>
  2.1 使用<code>input_size</code>, <code>output_size</code>和<code>block_num</code>来构造搜索空间。例如：传入SANAS的config可以指定为[('MobileNetV2BlockSpace', {'input_size': 224, 'output_size': 32, 'block_num': 10})]。<br>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
238 239 
  2.2 使用<code>block_mask</code>构造搜索空间。例如：传入SANAS的config可以指定为[('MobileNetV2BlockSpace', {'block_mask': [0, 1, 1, 1, 1, 0, 1, 0]})]。</li>
</ol>
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
240 241 242 243 244 
<h2 id="search-space">自定义搜索空间(search space)<a class="headerlink" href="#search-space" title="Permanent link">#</a></h2>
<p>自定义搜索空间类需要继承搜索空间基类并重写以下几部分：<br>
&emsp; 1. 初始化的tokens(<code>init_tokens</code>函数)，可以设置为自己想要的tokens列表, tokens列表中的每个数字指的是当前数字在相应的搜索列表中的索引。例如本示例中若tokens=[0, 3, 5]，则代表当前模型结构搜索到的通道数为[8, 40, 128]。<br>
&emsp; 2. token中每个数字的搜索列表长度(<code>range_table</code>函数)，tokens中每个token的索引范围。<br>
&emsp; 3. 根据token产生模型结构(<code>token2arch</code>函数)，根据搜索到的tokens列表产生模型结构。 <br></p>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
245 
<p>以新增reset block为例说明如何构造自己的search space。自定义的search space不能和已有的search space同名。</p>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
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 
<table class="codehilitetable"><tr><td class="linenos"><div class="linenodiv"><pre><span></span> 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
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
306 307 
61
62</pre></div></td><td class="code"><div class="codehilite"><pre><span></span><span class="c1">### 引入搜索空间基类函数和search space的注册类函数</span>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 
<span class="kn">from</span> <span class="nn">.search_space_base</span> <span class="kn">import</span> <span class="n">SearchSpaceBase</span>
<span class="kn">from</span> <span class="nn">.search_space_registry</span> <span class="kn">import</span> <span class="n">SEARCHSPACE</span>
<span class="kn">import</span> <span class="nn">numpy</span> <span class="kn">as</span> <span class="nn">np</span>

<span class="c1">### 需要调用注册函数把自定义搜索空间注册到space space中</span>
<span class="nd">@SEARCHSPACE.register</span>
<span class="c1">### 定义一个继承SearchSpaceBase基类的搜索空间的类函数</span>
<span class="k">class</span> <span class="nc">ResNetBlockSpace2</span><span class="p">(</span><span class="n">SearchSpaceBase</span><span class="p">):</span>
    <span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">input_size</span><span class="p">,</span> <span class="n">output_size</span><span class="p">,</span> <span class="n">block_num</span><span class="p">,</span> <span class="n">block_mask</span><span class="p">):</span>
        <span class="c1">### 定义一些实际想要搜索的内容，例如：通道数、每个卷积的重复次数、卷积核大小等等</span>
        <span class="c1">### self.filter_num 代表通道数的搜索列表</span>
        <span class="bp">self</span><span class="o">.</span><span class="n">filter_num</span> <span class="o">=</span> <span class="n">np</span><span class="o">.</span><span class="n">array</span><span class="p">([</span><span class="mi">8</span><span class="p">,</span> <span class="mi">16</span><span class="p">,</span> <span class="mi">32</span><span class="p">,</span> <span class="mi">40</span><span class="p">,</span> <span class="mi">64</span><span class="p">,</span> <span class="mi">128</span><span class="p">,</span> <span class="mi">256</span><span class="p">,</span> <span class="mi">512</span><span class="p">])</span>

    <span class="c1">### 定义初始化token，初始化token的长度根据传入的block_num或者block_mask的长度来得到的</span>
    <span class="k">def</span> <span class="nf">init_tokens</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
        <span class="k">return</span> <span class="p">[</span><span class="mi">0</span><span class="p">]</span> <span class="o">*</span> <span class="mi">3</span> <span class="o">*</span> <span class="nb">len</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">block_mask</span><span class="p">)</span>
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
325 
    <span class="c1">### 定义token的index的取值范围</span>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
326 327 328 
    <span class="k">def</span> <span class="nf">range_table</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
        <span class="k">return</span> <span class="p">[</span><span class="nb">len</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">filter_num</span><span class="p">)]</span> <span class="o">*</span> <span class="mi">3</span> <span class="o">*</span> <span class="nb">len</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">block_mask</span><span class="p">)</span>
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
329 
    <span class="c1">### 把token转换成模型结构</span>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
330 331 332 333 334 335 
    <span class="k">def</span> <span class="nf">token2arch</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">tokens</span><span class="o">=</span><span class="bp">None</span><span class="p">):</span>
        <span class="k">if</span> <span class="n">tokens</span> <span class="o">==</span> <span class="bp">None</span><span class="p">:</span>
            <span class="n">tokens</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">init_tokens</span><span class="p">()</span>

        <span class="bp">self</span><span class="o">.</span><span class="n">bottleneck_params_list</span> <span class="o">=</span> <span class="p">[]</span>
        <span class="k">for</span> <span class="n">i</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="nb">len</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">block_mask</span><span class="p">)):</span>
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
336 
            <span class="bp">self</span><span class="o">.</span><span class="n">bottleneck_params_list</span><span class="o">.</span><span class="n">append</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">filter_num</span><span class="p">[</span><span class="n">tokens</span><span class="p">[</span><span class="n">i</span> <span class="o">*</span> <span class="mi">3</span> <span class="o">+</span> <span class="mi">0</span><span class="p">]],</span>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
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 
                                               <span class="bp">self</span><span class="o">.</span><span class="n">filter_num</span><span class="p">[</span><span class="n">tokens</span><span class="p">[</span><span class="n">i</span> <span class="o">*</span> <span class="mi">3</span> <span class="o">+</span> <span class="mi">1</span><span class="p">]],</span>
                                               <span class="bp">self</span><span class="o">.</span><span class="n">filter_num</span><span class="p">[</span><span class="n">tokens</span><span class="p">[</span><span class="n">i</span> <span class="o">*</span> <span class="mi">3</span> <span class="o">+</span> <span class="mi">2</span><span class="p">]],</span>
                                               <span class="mi">2</span> <span class="k">if</span> <span class="bp">self</span><span class="o">.</span><span class="n">block_mask</span><span class="p">[</span><span class="n">i</span><span class="p">]</span> <span class="o">==</span> <span class="mi">1</span> <span class="k">else</span> <span class="mi">1</span><span class="p">)</span>

        <span class="k">def</span> <span class="nf">net_arch</span><span class="p">(</span><span class="nb">input</span><span class="p">):</span>
            <span class="k">for</span> <span class="n">i</span><span class="p">,</span> <span class="n">layer_setting</span> <span class="ow">in</span> <span class="nb">enumerate</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">bottleneck_params_list</span><span class="p">):</span>
                <span class="n">channel_num</span><span class="p">,</span> <span class="n">stride</span> <span class="o">=</span> <span class="n">layer_setting</span><span class="p">[:</span><span class="o">-</span><span class="mi">1</span><span class="p">],</span> <span class="n">layer_setting</span><span class="p">[</span><span class="o">-</span><span class="mi">1</span><span class="p">]</span>
                <span class="nb">input</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">_resnet_block</span><span class="p">(</span><span class="nb">input</span><span class="p">,</span> <span class="n">channel_num</span><span class="p">,</span> <span class="n">stride</span><span class="p">,</span> <span class="n">name</span><span class="o">=</span><span class="s1">&#39;resnet_layer{}&#39;</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">i</span><span class="o">+</span><span class="mi">1</span><span class="p">))</span>

            <span class="k">return</span> <span class="nb">input</span>

        <span class="k">return</span> <span class="n">net_arch</span>

    <span class="c1">### 构造具体block的操作</span>
    <span class="k">def</span> <span class="nf">_resnet_block</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="nb">input</span><span class="p">,</span> <span class="n">channel_num</span><span class="p">,</span> <span class="n">stride</span><span class="p">,</span> <span class="n">name</span><span class="o">=</span><span class="bp">None</span><span class="p">):</span>
        <span class="n">shortcut_conv</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">_shortcut</span><span class="p">(</span><span class="nb">input</span><span class="p">,</span> <span class="n">channel_num</span><span class="p">[</span><span class="mi">2</span><span class="p">],</span> <span class="n">stride</span><span class="p">,</span> <span class="n">name</span><span class="o">=</span><span class="n">name</span><span class="p">)</span>
        <span class="nb">input</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">_conv_bn_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span> <span class="n">num_filters</span><span class="o">=</span><span class="n">channel_num</span><span class="p">[</span><span class="mi">0</span><span class="p">],</span> <span class="n">filter_size</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span> <span class="n">act</span><span class="o">=</span><span class="s1">&#39;relu&#39;</span><span class="p">,</span> <span class="n">name</span><span class="o">=</span><span class="n">name</span> <span class="o">+</span> <span class="s1">&#39;_conv0&#39;</span><span class="p">)</span>
        <span class="nb">input</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">_conv_bn_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span> <span class="n">num_filters</span><span class="o">=</span><span class="n">channel_num</span><span class="p">[</span><span class="mi">1</span><span class="p">],</span> <span class="n">filter_size</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span> <span class="n">stride</span><span class="o">=</span><span class="n">stride</span><span class="p">,</span> <span class="n">act</span><span class="o">=</span><span class="s1">&#39;relu&#39;</span><span class="p">,</span> <span class="n">name</span><span class="o">=</span><span class="n">name</span> <span class="o">+</span> <span class="s1">&#39;_conv1&#39;</span><span class="p">)</span>
        <span class="nb">input</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">_conv_bn_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span> <span class="n">num_filters</span><span class="o">=</span><span class="n">channel_num</span><span class="p">[</span><span class="mi">2</span><span class="p">],</span> <span class="n">filter_size</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span> <span class="n">name</span><span class="o">=</span><span class="n">name</span> <span class="o">+</span> <span class="s1">&#39;_conv2&#39;</span><span class="p">)</span>
        <span class="k">return</span> <span class="n">fluid</span><span class="o">.</span><span class="n">layers</span><span class="o">.</span><span class="n">elementwise_add</span><span class="p">(</span><span class="n">x</span><span class="o">=</span><span class="n">shortcut_conv</span><span class="p">,</span> <span class="n">y</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span> <span class="n">axis</span><span class="o">=</span><span class="mi">0</span><span class="p">,</span> <span class="n">name</span><span class="o">=</span><span class="n">name</span><span class="o">+</span><span class="s1">&#39;_elementwise_add&#39;</span><span class="p">)</span>

    <span class="k">def</span> <span class="nf">_shortcut</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="nb">input</span><span class="p">,</span> <span class="n">channel_num</span><span class="p">,</span> <span class="n">stride</span><span class="p">,</span> <span class="n">name</span><span class="o">=</span><span class="bp">None</span><span class="p">):</span>
        <span class="n">channel_in</span> <span class="o">=</span> <span class="nb">input</span><span class="o">.</span><span class="n">shape</span><span class="p">[</span><span class="mi">1</span><span class="p">]</span>
        <span class="k">if</span> <span class="n">channel_in</span> <span class="o">!=</span> <span class="n">channel_num</span> <span class="ow">or</span> <span class="n">stride</span> <span class="o">!=</span> <span class="mi">1</span><span class="p">:</span>
            <span class="k">return</span> <span class="bp">self</span><span class="o">.</span><span class="n">conv_bn_layer</span><span class="p">(</span><span class="nb">input</span><span class="p">,</span> <span class="n">num_filters</span><span class="o">=</span><span class="n">channel_num</span><span class="p">,</span> <span class="n">filter_size</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span> <span class="n">stride</span><span class="o">=</span><span class="n">stride</span><span class="p">,</span> <span class="n">name</span><span class="o">=</span><span class="n">name</span><span class="o">+</span><span class="s1">&#39;_shortcut&#39;</span><span class="p">)</span>
        <span class="k">else</span><span class="p">:</span>
            <span class="k">return</span> <span class="nb">input</span>

    <span class="k">def</span> <span class="nf">_conv_bn_layer</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="nb">input</span><span class="p">,</span> <span class="n">num_filters</span><span class="p">,</span> <span class="n">filter_size</span><span class="p">,</span> <span class="n">stride</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span> <span class="n">padding</span><span class="o">=</span><span class="s1">&#39;SAME&#39;</span><span class="p">,</span> <span class="n">act</span><span class="o">=</span><span class="bp">None</span><span class="p">,</span> <span class="n">name</span><span class="o">=</span><span class="bp">None</span><span class="p">):</span>
        <span class="n">conv</span> <span class="o">=</span> <span class="n">fluid</span><span class="o">.</span><span class="n">layers</span><span class="o">.</span><span class="n">conv2d</span><span class="p">(</span><span class="nb">input</span><span class="p">,</span> <span class="n">num_filters</span><span class="p">,</span> <span class="n">filter_size</span><span class="p">,</span> <span class="n">stride</span><span class="p">,</span> <span class="n">name</span><span class="o">=</span><span class="n">name</span><span class="o">+</span><span class="s1">&#39;_conv&#39;</span><span class="p">)</span>
        <span class="n">bn</span> <span class="o">=</span> <span class="n">fluid</span><span class="o">.</span><span class="n">layers</span><span class="o">.</span><span class="n">batch_norm</span><span class="p">(</span><span class="n">conv</span><span class="p">,</span> <span class="n">act</span><span class="o">=</span><span class="n">act</span><span class="p">,</span> <span class="n">name</span><span class="o">=</span><span class="n">name</span><span class="o">+</span><span class="s1">&#39;_bn&#39;</span><span class="p">)</span>
        <span class="k">return</span> <span class="n">bn</span>
</pre></div>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
370 
</td></tr></table>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
371 372 373 374 375 
              
            </div>
          </div>
          <footer>
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
376 377 378 379 380 381 382 383 384 
    <div class="rst-footer-buttons" role="navigation" aria-label="footer navigation">
      
        <a href="../table_latency/" class="btn btn-neutral float-right" title="硬件延时评估表">Next <span class="icon icon-circle-arrow-right"></span></a>
      
      
        <a href="../api/nas_api/" class="btn btn-neutral" title="SA搜索"><span class="icon icon-circle-arrow-left"></span> Previous</a>
      
    </div>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 

  <hr/>

  <div role="contentinfo">
    <!-- Copyright etc -->
    
  </div>

  Built with <a href="http://www.mkdocs.org">MkDocs</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>

  <div class="rst-versions" role="note" style="cursor: pointer">
    <span class="rst-current-version" data-toggle="rst-current-version">
      
          <a href="https://github.com/PaddlePaddle/PaddleSlim/" class="fa fa-github" style="float: left; color: #fcfcfc"> GitHub</a>
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
409 410 411 412 
        <span><a href="../api/nas_api/" style="color: #fcfcfc;">&laquo; Previous</a></span>
      
      
        <span style="margin-left: 15px"><a href="../table_latency/" style="color: #fcfcfc">Next &raquo;</a></span>
B
Deployed 0f9c3ce with MkDocs version: 1.0.4  
baiyfbupt 已提交
413
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
414 415 
    </span>
</div>
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
416 417 418 
    <script>var base_url = '..';</script>
    <script src="../js/theme.js" defer></script>
      <script src="../mathjax-config.js" defer></script>
B
Deployed 0f9c3ce with MkDocs version: 1.0.4  
baiyfbupt 已提交
419 
      <script src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.0/MathJax.js?config=TeX-AMS-MML_HTMLorMML" defer></script>
B
Deployed ab60054 with MkDocs version: 1.0.4  
baiyfbupt 已提交
420 
      <script src="../search/main.js" defer></script>
B
Deployed 09f99c9 with MkDocs version: 1.0.4  
baiyfbupt 已提交
421 422 423 

</body>
</html>