Skip to content

P
Paddle

BaiXuePrincess / Paddle
与 Fork 源项目一致

Fork自 PaddlePaddle / Paddle

通知 1
Star 1
Fork 0
P
Paddle
提交 8f5e20e7 编写于 作者: T Travis CI
浏览文件
操作

Deploy to GitHub Pages: cebdb667

上级 bc4f5adb
隐藏空白更改
内联 并排
Showing with 37 addition and 532 deletion
doc/genindex.html
浏览文件 @ 8f5e20e7
......@@ -1823,7 +1823,7 @@ var _hmt = _hmt || [];
</dt>
<dt><a href="source/cuda/rnn/rnn.html#_CPPv222hl_sequence2batch_copyP4realP4realPKiiib">hl_sequence2batch_copy (C++ function)</a>
<dt><a href="source/cuda/rnn/rnn.html#_CPPv222hl_sequence2batch_copyP4realP4realPiiib">hl_sequence2batch_copy (C++ function)</a>
</dt>
......@@ -5247,7 +5247,7 @@ var _hmt = _hmt || [];
</dt>
<dt><a href="source/math/matrix/matrix.html#_CPPv2N6paddle9CpuMatrix14copyByRowIndexER6MatrixRK7IVector">paddle::CpuMatrix::copyByRowIndex (C++ function)</a>
<dt><a href="source/math/matrix/matrix.html#_CPPv2N6paddle9CpuMatrix14copyByRowIndexER6MatrixR7IVector">paddle::CpuMatrix::copyByRowIndex (C++ function)</a>
</dt>
......@@ -6955,6 +6955,10 @@ var _hmt = _hmt || [];
</dt>
<dt><a href="source/gserver/layers/layer.html#_CPPv2N6paddle11ExpandLayer19cpuExpandStartsPos_E">paddle::ExpandLayer::cpuExpandStartsPos_ (C++ member)</a>
</dt>
<dt><a href="source/gserver/layers/layer.html#_CPPv2N6paddle11ExpandLayer11ExpandLayerERK11LayerConfig">paddle::ExpandLayer::ExpandLayer (C++ function)</a>
</dt>
......@@ -7355,7 +7359,7 @@ var _hmt = _hmt || [];
</dt>
<dt><a href="source/math/matrix/matrix.html#_CPPv2N6paddle9GpuMatrix14copyByRowIndexER6MatrixRK7IVector">paddle::GpuMatrix::copyByRowIndex (C++ function)</a>
<dt><a href="source/math/matrix/matrix.html#_CPPv2N6paddle9GpuMatrix14copyByRowIndexER6MatrixR7IVector">paddle::GpuMatrix::copyByRowIndex (C++ function)</a>
</dt>
......@@ -9335,7 +9339,7 @@ var _hmt = _hmt || [];
</dt>
<dt><a href="source/math/matrix/matrix.html#_CPPv2N6paddle6Matrix14copyByRowIndexER6MatrixRK7IVector">paddle::Matrix::copyByRowIndex (C++ function)</a>
<dt><a href="source/math/matrix/matrix.html#_CPPv2N6paddle6Matrix14copyByRowIndexER6MatrixR7IVector">paddle::Matrix::copyByRowIndex (C++ function)</a>
</dt>
......@@ -14461,6 +14465,10 @@ var _hmt = _hmt || [];
</dt>
<dt><a href="source/gserver/layers/layer.html#_CPPv2N6paddle25SequenceScatterAgentLayer17cpuInputStartPos_E">paddle::SequenceScatterAgentLayer::cpuInputStartPos_ (C++ member)</a>
</dt>
<dt><a href="source/gserver/layers/layer.html#_CPPv2N6paddle25SequenceScatterAgentLayer7forwardE8PassType">paddle::SequenceScatterAgentLayer::forward (C++ function)</a>
</dt>
......
doc/objects.inv
浏览文件 @ 8f5e20e7
无法预览此类型文件
doc/searchindex.js
浏览文件 @ 8f5e20e7
因为 它太大了无法显示 source diff 。你可以改为 查看blob。
doc/source/cuda/rnn/rnn.html
浏览文件 @ 8f5e20e7
......@@ -1013,8 +1013,8 @@ var _hmt = _hmt || [];
</dd></dl>
<dl class="function">
<dt id="_CPPv222hl_sequence2batch_copyP4realP4realPKiiib">
<span id="hl_sequence2batch_copy__realP.realP.iCP.i.i.b"></span><span class="target" id="paddlehl__sequence_8h_1a13d7f834880527645555849e05278745"></span>void <code class="descname">hl_sequence2batch_copy</code><span class="sig-paren">(</span>real *<em>batch</em>, real *<em>sequence</em>, <em class="property">const</em> int *<em>batchIndex</em>, int <em>seqWidth</em>, int <em>batchCount</em>, bool <em>seq2batch</em><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv222hl_sequence2batch_copyP4realP4realPKiiib" title="Permalink to this definition"></a></dt>
<dt id="_CPPv222hl_sequence2batch_copyP4realP4realPiiib">
<span id="hl_sequence2batch_copy__realP.realP.iP.i.i.b"></span><span class="target" id="paddlehl__sequence_8h_1a6e0b30bd2703b8232ac1d70022306a6a"></span>void <code class="descname">hl_sequence2batch_copy</code><span class="sig-paren">(</span>real *<em>batch</em>, real *<em>sequence</em>, int *<em>batchIndex</em>, int <em>seqWidth</em>, int <em>batchCount</em>, bool <em>seq2batch</em><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv222hl_sequence2batch_copyP4realP4realPiiib" title="Permalink to this definition"></a></dt>
<dd><p>Memory copy from sequence to batch. </p>
<p>if seq2batch == true</p>
<p>copy from sequence to batch: batch[i] = sequence[batchIndex[i]].</p>
......
doc/source/gserver/layers/layer.html
浏览文件 @ 8f5e20e7
......@@ -4623,9 +4623,14 @@ The config file api if gru_step_layer. <dl class="docutils">
</div>
<div class="breathe-sectiondef container">
<p class="breathe-sectiondef-title rubric">Protected Attributes</p>
<dl class="member">
<dt id="_CPPv2N6paddle25SequenceScatterAgentLayer17cpuInputStartPos_E">
<span id="paddle::SequenceScatterAgentLayer::cpuInputStartPos___IVectorPtr"></span><span class="target" id="paddleclasspaddle_1_1SequenceScatterAgentLayer_1a0fd54096dd1552a7e42b9ca1cbbf50ae"></span><a class="reference internal" href="../../math/matrix/matrix.html#_CPPv2N6paddle10IVectorPtrE" title="paddle::IVectorPtr">IVectorPtr</a> <code class="descname">cpuInputStartPos_</code><a class="headerlink" href="#_CPPv2N6paddle25SequenceScatterAgentLayer17cpuInputStartPos_E" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<dl class="member">
<dt id="_CPPv2N6paddle25SequenceScatterAgentLayer14inputStartPos_E">
<span id="paddle::SequenceScatterAgentLayer::inputStartPos___ICpuGpuVectorPtr"></span><span class="target" id="paddleclasspaddle_1_1SequenceScatterAgentLayer_1acfcf479183ea7b96c05968d2a5ce414e"></span><a class="reference internal" href="../../math/matrix/matrix.html#_CPPv2N6paddle16ICpuGpuVectorPtrE" title="paddle::ICpuGpuVectorPtr">ICpuGpuVectorPtr</a> <code class="descname">inputStartPos_</code><a class="headerlink" href="#_CPPv2N6paddle25SequenceScatterAgentLayer14inputStartPos_E" title="Permalink to this definition">¶</a></dt>
<span id="paddle::SequenceScatterAgentLayer::inputStartPos___IVectorPtr"></span><span class="target" id="paddleclasspaddle_1_1SequenceScatterAgentLayer_1a3026312851c50c9952446a7440a07870"></span><a class="reference internal" href="../../math/matrix/matrix.html#_CPPv2N6paddle10IVectorPtrE" title="paddle::IVectorPtr">IVectorPtr</a> <code class="descname">inputStartPos_</code><a class="headerlink" href="#_CPPv2N6paddle25SequenceScatterAgentLayer14inputStartPos_E" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
</div>
......@@ -5743,11 +5748,17 @@ sequence is one) to sequence data.&#8221;</p>
</dd></dl>
<dl class="member">
<dt id="_CPPv2N6paddle11ExpandLayer16expandStartsPos_E">
<span id="paddle::ExpandLayer::expandStartsPos___ICpuGpuVectorPtr"></span><span class="target" id="paddleclasspaddle_1_1ExpandLayer_1a4243425f33452a5d1ac468d257ce111a"></span><a class="reference internal" href="../../math/matrix/matrix.html#_CPPv2N6paddle16ICpuGpuVectorPtrE" title="paddle::ICpuGpuVectorPtr">ICpuGpuVectorPtr</a> <code class="descname">expandStartsPos_</code><a class="headerlink" href="#_CPPv2N6paddle11ExpandLayer16expandStartsPos_E" title="Permalink to this definition">¶</a></dt>
<dt id="_CPPv2N6paddle11ExpandLayer19cpuExpandStartsPos_E">
<span id="paddle::ExpandLayer::cpuExpandStartsPos___IVectorPtr"></span><span class="target" id="paddleclasspaddle_1_1ExpandLayer_1af081e0e3d584c24737be9e9f431914b8"></span><a class="reference internal" href="../../math/matrix/matrix.html#_CPPv2N6paddle10IVectorPtrE" title="paddle::IVectorPtr">IVectorPtr</a> <code class="descname">cpuExpandStartsPos_</code><a class="headerlink" href="#_CPPv2N6paddle11ExpandLayer19cpuExpandStartsPos_E" title="Permalink to this definition">¶</a></dt>
<dd><p>expanded sequenceStartPositions or subSequenceStartPositions of input[1] </p>
</dd></dl>
<dl class="member">
<dt id="_CPPv2N6paddle11ExpandLayer16expandStartsPos_E">
<span id="paddle::ExpandLayer::expandStartsPos___IVectorPtr"></span><span class="target" id="paddleclasspaddle_1_1ExpandLayer_1a6528e39055e5de7245e6fcdd83a96c74"></span><a class="reference internal" href="../../math/matrix/matrix.html#_CPPv2N6paddle10IVectorPtrE" title="paddle::IVectorPtr">IVectorPtr</a> <code class="descname">expandStartsPos_</code><a class="headerlink" href="#_CPPv2N6paddle11ExpandLayer16expandStartsPos_E" title="Permalink to this definition">¶</a></dt>
<dd><p>point to cpuExpandStartsPos_ when useGpu_ is false, copy from cpuExpandStartsPos_ when useGpu_ is true </p>
</dd></dl>
</div>
</dd></dl>
......
doc/source/math/matrix/matrix.html
浏览文件 @ 8f5e20e7
......@@ -1756,8 +1756,8 @@ virtual <span class="target" id="paddleclasspaddle_1_1Matrix_1aa8a2ffb8e06ea97ce
</dd></dl>
<dl class="function">
<dt id="_CPPv2N6paddle6Matrix14copyByRowIndexER6MatrixRK7IVector">
<span id="paddle::Matrix::copyByRowIndex__MatrixR.IVectorCR"></span>virtual <span class="target" id="paddleclasspaddle_1_1Matrix_1ae0eb1e4febf16bff85f1119ba18dedb0"></span>void <code class="descname">copyByRowIndex</code><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv2N6paddle6MatrixE" title="paddle::Matrix">Matrix</a> &amp;<em>b</em>, <em class="property">const</em> <a class="reference internal" href="#_CPPv2N6paddle7IVectorE" title="paddle::IVector">IVector</a> &amp;<em>rowIndex</em><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv2N6paddle6Matrix14copyByRowIndexER6MatrixRK7IVector" title="Permalink to this definition">¶</a></dt>
<dt id="_CPPv2N6paddle6Matrix14copyByRowIndexER6MatrixR7IVector">
<span id="paddle::Matrix::copyByRowIndex__MatrixR.IVectorR"></span>virtual <span class="target" id="paddleclasspaddle_1_1Matrix_1a4f59ce1c02e1516a53b1dd5247d8d356"></span>void <code class="descname">copyByRowIndex</code><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv2N6paddle6MatrixE" title="paddle::Matrix">Matrix</a> &amp;<em>b</em>, <a class="reference internal" href="#_CPPv2N6paddle7IVectorE" title="paddle::IVector">IVector</a> &amp;<em>rowIndex</em><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv2N6paddle6Matrix14copyByRowIndexER6MatrixR7IVector" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<dl class="function">
......@@ -2551,8 +2551,8 @@ where bit(i, j) = ((codes(i) + numClasses) &amp; 2^j) ? 1 : 0
</dd></dl>
<dl class="function">
<dt id="_CPPv2N6paddle9GpuMatrix14copyByRowIndexER6MatrixRK7IVector">
<span id="paddle::GpuMatrix::copyByRowIndex__MatrixR.IVectorCR"></span>virtual <span class="target" id="paddleclasspaddle_1_1GpuMatrix_1a7d757b48a3fcfcd594818c48a88b357c"></span>void <code class="descname">copyByRowIndex</code><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv2N6paddle6MatrixE" title="paddle::Matrix">Matrix</a> &amp;<em>b</em>, <em class="property">const</em> <a class="reference internal" href="#_CPPv2N6paddle7IVectorE" title="paddle::IVector">IVector</a> &amp;<em>rowIndex</em><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv2N6paddle9GpuMatrix14copyByRowIndexER6MatrixRK7IVector" title="Permalink to this definition">¶</a></dt>
<dt id="_CPPv2N6paddle9GpuMatrix14copyByRowIndexER6MatrixR7IVector">
<span id="paddle::GpuMatrix::copyByRowIndex__MatrixR.IVectorR"></span>virtual <span class="target" id="paddleclasspaddle_1_1GpuMatrix_1a46f305110a92a4fbdc98faaba0a35951"></span>void <code class="descname">copyByRowIndex</code><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv2N6paddle6MatrixE" title="paddle::Matrix">Matrix</a> &amp;<em>b</em>, <a class="reference internal" href="#_CPPv2N6paddle7IVectorE" title="paddle::IVector">IVector</a> &amp;<em>rowIndex</em><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv2N6paddle9GpuMatrix14copyByRowIndexER6MatrixR7IVector" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<dl class="function">
......@@ -3111,8 +3111,8 @@ where bit(i, j) = ((codes(i) + numClasses) &amp; 2^j) ? 1 : 0
<dd></dd></dl>
<dl class="function">
<dt id="_CPPv2N6paddle9CpuMatrix14copyByRowIndexER6MatrixRK7IVector">
<span id="paddle::CpuMatrix::copyByRowIndex__MatrixR.IVectorCR"></span>virtual <span class="target" id="paddleclasspaddle_1_1CpuMatrix_1a8bfc00af2e2e972193b762434e26d979"></span>void <code class="descname">copyByRowIndex</code><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv2N6paddle6MatrixE" title="paddle::Matrix">Matrix</a> &amp;<em>b</em>, <em class="property">const</em> <a class="reference internal" href="#_CPPv2N6paddle7IVectorE" title="paddle::IVector">IVector</a> &amp;<em>rowIndex</em><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv2N6paddle9CpuMatrix14copyByRowIndexER6MatrixRK7IVector" title="Permalink to this definition">¶</a></dt>
<dt id="_CPPv2N6paddle9CpuMatrix14copyByRowIndexER6MatrixR7IVector">
<span id="paddle::CpuMatrix::copyByRowIndex__MatrixR.IVectorR"></span>virtual <span class="target" id="paddleclasspaddle_1_1CpuMatrix_1a24837d660ae6691384313b5d399ad5fe"></span>void <code class="descname">copyByRowIndex</code><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv2N6paddle6MatrixE" title="paddle::Matrix">Matrix</a> &amp;<em>b</em>, <a class="reference internal" href="#_CPPv2N6paddle7IVectorE" title="paddle::IVector">IVector</a> &amp;<em>rowIndex</em><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv2N6paddle9CpuMatrix14copyByRowIndexER6MatrixR7IVector" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<dl class="function">
......
doc_cn/.buildinfo
浏览文件 @ 8f5e20e7
# Sphinx build info version 1
# This file hashes the configuration used when building these files. When it is not found, a full rebuild will be done.
config: 8f9e3b6337374f468cc7e48534c4662a
config: 70a318b9e7a63a79aedc16f559247671
tags: 645f666f9bcd5a90fca523b33c5a78b7
doc_cn/_sources/faq/index.txt 已删除 100644 → 0
浏览文件 @ bc4f5adb
####################
PaddlePaddle常见问题
####################
.. contents::
1. 如何减少PaddlePaddle的内存占用
---------------------------------
神经网络的训练本身是一个非常消耗内存和显存的工作。经常会消耗数十G的内存和数G的显存。
PaddlePaddle的内存占用主要分为如下几个方面\:
* DataProvider缓冲池内存 (只针对内存)
* 神经元激活内存 （针对内存和显存）
* 参数内存 (针对内存和显存)
* 其他内存杂项
这其中，其他内存杂项是指PaddlePaddle本身所用的一些内存，包括字符串分配，临时变量等等，
这些内存就不考虑如何缩减了。
其他的内存的减少方法依次为
减少DataProvider缓冲池内存
++++++++++++++++++++++++++
PyDataProvider使用的是异步加载，同时在内存里直接随即选取数据来做Shuffle。即
.. graphviz::
digraph {
rankdir=LR;
数据文件 -> 内存池 -> PaddlePaddle训练
}
所以，减小这个内存池即可减小内存占用，同时也可以加速开始训练前数据载入的过程。但是，这
个内存池实际上决定了shuffle的粒度。所以，如果将这个内存池减小，又要保证数据是随机的，
那么最好将数据文件在每次读取之前做一次shuffle。可能的代码为
.. literalinclude:: reduce_min_pool_size.py
这样做可以极大的减少内存占用，并且可能会加速训练过程。 详细文档参考 `这里
<../ui/data_provider/pydataprovider2.html#provider>`_ 。
神经元激活内存
++++++++++++++
神经网络在训练的时候，会对每一个激活暂存一些数据，包括激活，參差等等。
在反向传递的时候，这些数据会被用来更新参数。这些数据使用的内存主要和两个参数有关系，
一是batch size，另一个是每条序列(Sequence)长度。所以，其实也是和每个mini-batch中包含
的时间步信息成正比。
所以，做法可以有两种。他们是
* 减小batch size。 即在网络配置中 :code:`settings(batch_size=1000)` 设置成一个小一些的值。但是batch size本身是神经网络的超参数，减小batch size可能会对训练结果产生影响。
* 减小序列的长度，或者直接扔掉非常长的序列。比如，一个数据集大部分序列长度是100-200,
但是突然有一个10000长的序列，就很容易导致内存超限。特别是在LSTM等RNN中。
参数内存
++++++++
PaddlePaddle支持非常多的优化算法(Optimizer)，不同的优化算法需要使用不同大小的内存。
例如如果使用 :code:`adadelta` 算法，则需要使用参数规模大约5倍的内存。 如果参数保存下来的
文件为 :code:`100M`， 那么该优化算法至少需要 :code:`500M` 的内存。
可以考虑使用一些优化算法，例如 :code:`momentum`。
2. 如何加速PaddlePaddle的训练速度
---------------------------------
PaddlePaddle是神经网络训练平台，加速PaddlePaddle训练有如下几个方面\：
* 减少数据载入的耗时
* 加速训练速度
* 利用更多的计算资源
减少数据载入的耗时
++++++++++++++++++
使用 :code:`pydataprovider`时，可以减少缓存池的大小，同时设置内存缓存功能，即可以极大的加速数据载入流程。
:code:`DataProvider` 缓存池的减小，和之前减小通过减小缓存池来减小内存占用的原理一致。
.. literalinclude:: reduce_min_pool_size.py
同时 :code:`@provider` 接口有一个 :code:`cache` 参数来控制缓存方法，将其设置成 :code:`CacheType.CACHE_PASS_IN_MEM` 的话，会将第一个 :code:`pass` (过完所有训练数据即为一个pass)生成的数据缓存在内存里，在之后的 :code:`pass` 中，不会再从 :code:`python` 端读取数据，而是直接从内存的缓存里读取数据。这也会极大减少数据读入的耗时。
加速训练速度
++++++++++++
PaddlePaddle支持Sparse的训练，sparse训练需要训练特征是 :code:`sparse_binary_vector` 、 :code:`sparse_vector` 、或者 :code:`integer_value` 的任一一种。同时，与这个训练数据交互的Layer，需要将其Parameter设置成 sparse 更新模式，即设置 :code:`sparse_update=True`
这里使用简单的 :code:`word2vec` 训练语言模型距离，具体使用方法为\:
使用一个词前两个词和后两个词，来预测这个中间的词。这个任务的DataProvider为\:
.. literalinclude:: word2vec_dataprovider.py
这个任务的配置为\:
.. literalinclude:: word2vec_config.py
更多关于sparse训练的内容请参考 `sparse训练的文档 <TBD>`_
利用更多的计算资源
++++++++++++++++++
利用更多的计算资源可以分为一下几个方式来进行\:
* 单机CPU训练
* 使用多线程训练。设置命令行参数 :code:`trainer_count`，即可以设置参与训练的线程数量。使用方法为 :code:`paddle train --trainer_count=4`
* 单机GPU训练
* 使用显卡训练。设置命令行参数 :code:`use_gpu`。 使用方法为 :code:`paddle train --use_gpu=true`
* 使用多块显卡训练。设置命令行参数 :code:`use_gpu` 和 :code:`trainer_count`。使用 :code:`--use_gpu=True` 开启GPU训练，使用 :code:`trainer_count` 指定显卡数量。使用方法为 :code:`paddle train --use_gpu=true --trainer_count=4`
* 多机训练
* 使用多机训练的方法也比较简单，需要先在每个节点启动 :code:`paddle pserver`，在使用 :code:`paddle train --pservers=192.168.100.1,192.168.100.2` 来指定每个pserver的ip地址
* 具体的多机训练方法参考 `多机训练 <TBD>`_ 文档。
3. 遇到“非法指令”或者是“illegal instruction”
--------------------------------------------
paddle在进行计算的时候为了提升计算性能，使用了avx指令。部分老的cpu型号无法支持这样的指令。通常来说执行下grep avx /proc/cpuinfo看看是否有输出即可知道是否支持。（另：用此方法部分虚拟机可能检测到支持avx指令但是实际运行会挂掉，请当成是不支持，看下面的解决方案）
解决办法是\:
* 使用 NO_AVX的 `安装包 <../build_and_install/index.html>`_ 或者 `Docker image <../build_and_install/install/docker_install.html>`_
* 或者，使用 :code:`-DWITH_AVX=OFF` 重新编译PaddlePaddle。
4. 如何选择SGD算法的学习率
--------------------------
在采用sgd/async_sgd进行训练时，一个重要的问题是选择正确的learning_rate。如果learning_rate太大，那么训练有可能不收敛，如果learning_rate太小，那么收敛可能很慢，导致训练时间过长。
通常做法是从一个比较大的learning_rate开始试，如果不收敛，那减少学习率10倍继续试验，直到训练收敛为止。那么如何判断训练不收敛呢？可以估计出如果模型采用不变的输出最小的cost0是多少。
如果训练过程的的cost明显高于这个常数输出的cost，那么我们可以判断为训练不收敛。举一个例子，假如我们是三分类问题，采用multi-class-cross-entropy作为cost，数据中0,1,2三类的比例为 :code:`0.2, 0.5, 0.3` , 那么常数输出所能达到的最小cost是 :code:`-(0.2*log(0.2)+0.5*log(0.5)+0.3*log(0.3))=1.03` 。如果训练一个pass（或者更早）后，cost还大于这个数，那么可以认为训练不收敛，应该降低学习率。
5. 如何初始化参数
-----------------
默认情况下，PaddlePaddle使用均值0，标准差为 :math:`\frac{1}{\sqrt{d}}` 来初始化参数。其中 :math:`d` 为参数矩阵的宽度。这种初始化方式在一般情况下不会产生很差的结果。如果用户想要自定义初始化方式，PaddlePaddle目前提供两种参数初始化的方式\:
* 高斯分布。将 :code:`param_attr` 设置成 :code:`param_attr=ParamAttr(initial_mean=0.0, initial_std=1.0)`
* 均匀分布。将 :code:`param_attr` 设置成 :code:`param_attr=ParamAttr(initial_max=1.0, initial_min=-1.0)`
比如设置一个全连接层的参数初始化方式和bias初始化方式，可以使用如下代码。
.. code-block:: python
hidden = fc_layer(input=ipt, param_attr=ParamAttr(initial_max=1.0, initial_min=-1.0),
bias_attr=ParamAttr(initial_mean=1.0, initial_std=0.0))
上述代码将bias全部初始化为1.0, 同时将参数初始化为 :code:`[1.0, -1.0]` 的均匀分布。
6. 如何共享参数
---------------
PaddlePaddle的参数使用名字 :code:`name` 作为参数的ID，相同名字的参数，会共享参数。设置参数的名字，可以使用 :code:`ParamAttr(name="YOUR_PARAM_NAME")` 来设置。更方便的设置方式，是想要共享的参数使用同样的 :code:`ParamAttr` 对象。
简单的全连接网络，参数共享的配置示例为\:
.. literalinclude:: ../../python/paddle/trainer_config_helpers/tests/configs/shared_fc.py
这里 :code:`hidden_a` 和 :code:`hidden_b` 使用了同样的parameter和bias。并且softmax层的两个输入也使用了同样的参数 :code:`softmax_param`。
doc_cn/_sources/index.txt
浏览文件 @ 8f5e20e7
......@@ -3,7 +3,6 @@ PaddlePaddle文档
使用指南
--------
* `快速入门 <demo/quick_start/index.html>`_
* `编译与安装 <build_and_install/index.html>`_
* `用户接口 <ui/index.html>`_
......@@ -17,13 +16,7 @@ PaddlePaddle文档
算法教程
--------
* `Recurrent Group教程 <algorithm/rnn/rnn-tutorial.html>`_
* `单层RNN示例 <../doc/algorithm/rnn/rnn.html>`_
* `双层RNN示例 <algorithm/rnn/hierarchical-rnn.html>`_
* `支持双层序列作为输入的Layer <algorithm/rnn/hierarchical-layer.html>`_
常见问题
--------
* `常见问题 <faq/index.html>`_
doc_cn/faq/index.html 已删除 100644 → 0
浏览文件 @ bc4f5adb
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<title>PaddlePaddle常见问题 &#8212; PaddlePaddle documentation</title>
<link rel="stylesheet" href="../_static/classic.css" type="text/css" />
<link rel="stylesheet" href="../_static/pygments.css" type="text/css" />
<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>
<link rel="index" title="Index" href="../genindex.html" />
<link rel="search" title="Search" href="../search.html" />
<link rel="top" title="PaddlePaddle documentation" href="../index.html" />
<script>
var _hmt = _hmt || [];
(function() {
var hm = document.createElement("script");
hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
var s = document.getElementsByTagName("script")[0];
s.parentNode.insertBefore(hm, s);
})();
</script>
</head>
<body role="document">
<div class="related" role="navigation" aria-label="related navigation">
<h3>Navigation</h3>
<ul>
<li class="right" style="margin-right: 10px">
<a href="../genindex.html" title="General Index"
accesskey="I">index</a></li>
<li class="nav-item nav-item-0"><a href="../index.html">PaddlePaddle documentation</a> &#187;</li>
</ul>
</div>
<div class="document">
<div class="documentwrapper">
<div class="bodywrapper">
<div class="body" role="main">
<div class="section" id="paddlepaddle">
<h1><a class="toc-backref" href="#id13">PaddlePaddle常见问题</a><a class="headerlink" href="#paddlepaddle" title="Permalink to this headline"></a></h1>
<div class="contents topic" id="contents">
<p class="topic-title first">Contents</p>
<ul class="simple">
<li><a class="reference internal" href="#paddlepaddle" id="id13">PaddlePaddle常见问题</a><ul>
<li><a class="reference internal" href="#id1" id="id14">1. 如何减少PaddlePaddle的内存占用</a><ul>
<li><a class="reference internal" href="#dataprovider" id="id15">减少DataProvider缓冲池内存</a></li>
<li><a class="reference internal" href="#id3" id="id16">神经元激活内存</a></li>
<li><a class="reference internal" href="#id4" id="id17">参数内存</a></li>
</ul>
</li>
<li><a class="reference internal" href="#id5" id="id18">2. 如何加速PaddlePaddle的训练速度</a><ul>
<li><a class="reference internal" href="#id6" id="id19">减少数据载入的耗时</a></li>
<li><a class="reference internal" href="#id7" id="id20">加速训练速度</a></li>
<li><a class="reference internal" href="#id8" id="id21">利用更多的计算资源</a></li>
</ul>
</li>
<li><a class="reference internal" href="#illegal-instruction" id="id22">3. 遇到“非法指令”或者是“illegal instruction”</a></li>
<li><a class="reference internal" href="#sgd" id="id23">4. 如何选择SGD算法的学习率</a></li>
<li><a class="reference internal" href="#id11" id="id24">5. 如何初始化参数</a></li>
<li><a class="reference internal" href="#id12" id="id25">6. 如何共享参数</a></li>
</ul>
</li>
</ul>
</div>
<div class="section" id="id1">
<h2><a class="toc-backref" href="#id14">1. 如何减少PaddlePaddle的内存占用</a><a class="headerlink" href="#id1" title="Permalink to this headline"></a></h2>
<p>神经网络的训练本身是一个非常消耗内存和显存的工作。经常会消耗数十G的内存和数G的显存。
PaddlePaddle的内存占用主要分为如下几个方面:</p>
<ul class="simple">
<li>DataProvider缓冲池内存 (只针对内存)</li>
<li>神经元激活内存 （针对内存和显存）</li>
<li>参数内存 (针对内存和显存)</li>
<li>其他内存杂项</li>
</ul>
<p>这其中，其他内存杂项是指PaddlePaddle本身所用的一些内存，包括字符串分配，临时变量等等，
这些内存就不考虑如何缩减了。</p>
<p>其他的内存的减少方法依次为</p>
<div class="section" id="dataprovider">
<h3><a class="toc-backref" href="#id15">减少DataProvider缓冲池内存</a><a class="headerlink" href="#dataprovider" title="Permalink to this headline"></a></h3>
<p>PyDataProvider使用的是异步加载，同时在内存里直接随即选取数据来做Shuffle。即</p>
<img src="../_images/graphviz-9be6aad37f57c60f4b971dde0ef44ce27179cf9a.png" alt="digraph {
rankdir=LR;
数据文件 -&gt; 内存池 -&gt; PaddlePaddle训练
}" />
<p>所以，减小这个内存池即可减小内存占用，同时也可以加速开始训练前数据载入的过程。但是，这
个内存池实际上决定了shuffle的粒度。所以，如果将这个内存池减小，又要保证数据是随机的，
那么最好将数据文件在每次读取之前做一次shuffle。可能的代码为</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="nd">@provider</span><span class="p">(</span><span class="n">min_pool_size</span><span class="o">=</span><span class="mi">0</span><span class="p">,</span> <span class="o">...</span><span class="p">)</span>
<span class="k">def</span> <span class="nf">process</span><span class="p">(</span><span class="n">settings</span><span class="p">,</span> <span class="n">filename</span><span class="p">):</span>
<span class="n">os</span><span class="o">.</span><span class="n">system</span><span class="p">(</span><span class="s1">&#39;shuf </span><span class="si">%s</span><span class="s1"> &gt; </span><span class="si">%s</span><span class="s1">.shuf&#39;</span> <span class="o">%</span> <span class="p">(</span><span class="n">filename</span><span class="p">,</span> <span class="n">filename</span><span class="p">))</span> <span class="c1"># shuffle before.</span>
<span class="k">with</span> <span class="nb">open</span><span class="p">(</span><span class="s1">&#39;</span><span class="si">%s</span><span class="s1">.shuf&#39;</span> <span class="o">%</span> <span class="n">filename</span><span class="p">,</span> <span class="s1">&#39;r&#39;</span><span class="p">)</span> <span class="k">as</span> <span class="n">f</span><span class="p">:</span>
<span class="k">for</span> <span class="n">line</span> <span class="ow">in</span> <span class="n">f</span><span class="p">:</span>
<span class="k">yield</span> <span class="n">get_sample_from_line</span><span class="p">(</span><span class="n">line</span><span class="p">)</span>
</pre></div>
</div>
<p>这样做可以极大的减少内存占用，并且可能会加速训练过程。 详细文档参考 <a class="reference external" href="../ui/data_provider/pydataprovider2.html#provider">这里</a> 。</p>
</div>
<div class="section" id="id3">
<h3><a class="toc-backref" href="#id16">神经元激活内存</a><a class="headerlink" href="#id3" title="Permalink to this headline"></a></h3>
<p>神经网络在训练的时候，会对每一个激活暂存一些数据，包括激活，參差等等。
在反向传递的时候，这些数据会被用来更新参数。这些数据使用的内存主要和两个参数有关系，
一是batch size，另一个是每条序列(Sequence)长度。所以，其实也是和每个mini-batch中包含
的时间步信息成正比。</p>
<p>所以，做法可以有两种。他们是</p>
<ul class="simple">
<li>减小batch size。 即在网络配置中 <code class="code docutils literal"><span class="pre">settings(batch_size=1000)</span></code> 设置成一个小一些的值。但是batch size本身是神经网络的超参数，减小batch size可能会对训练结果产生影响。</li>
<li>减小序列的长度，或者直接扔掉非常长的序列。比如，一个数据集大部分序列长度是100-200,
但是突然有一个10000长的序列，就很容易导致内存超限。特别是在LSTM等RNN中。</li>
</ul>
</div>
<div class="section" id="id4">
<h3><a class="toc-backref" href="#id17">参数内存</a><a class="headerlink" href="#id4" title="Permalink to this headline"></a></h3>
<p>PaddlePaddle支持非常多的优化算法(Optimizer)，不同的优化算法需要使用不同大小的内存。
例如如果使用 <code class="code docutils literal"><span class="pre">adadelta</span></code> 算法，则需要使用参数规模大约5倍的内存。 如果参数保存下来的
文件为 <code class="code docutils literal"><span class="pre">100M</span></code>， 那么该优化算法至少需要 <code class="code docutils literal"><span class="pre">500M</span></code> 的内存。</p>
<p>可以考虑使用一些优化算法，例如 <code class="code docutils literal"><span class="pre">momentum</span></code>。</p>
</div>
</div>
<div class="section" id="id5">
<h2><a class="toc-backref" href="#id18">2. 如何加速PaddlePaddle的训练速度</a><a class="headerlink" href="#id5" title="Permalink to this headline"></a></h2>
<p>PaddlePaddle是神经网络训练平台，加速PaddlePaddle训练有如下几个方面：</p>
<ul class="simple">
<li>减少数据载入的耗时</li>
<li>加速训练速度</li>
<li>利用更多的计算资源</li>
</ul>
<div class="section" id="id6">
<h3><a class="toc-backref" href="#id19">减少数据载入的耗时</a><a class="headerlink" href="#id6" title="Permalink to this headline"></a></h3>
<p>使用 <code class="code docutils literal"><span class="pre">pydataprovider`时，可以减少缓存池的大小，同时设置内存缓存功能，即可以极大的加速数据载入流程。</span>
<span class="pre">:code:`DataProvider</span></code> 缓存池的减小，和之前减小通过减小缓存池来减小内存占用的原理一致。</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="nd">@provider</span><span class="p">(</span><span class="n">min_pool_size</span><span class="o">=</span><span class="mi">0</span><span class="p">,</span> <span class="o">...</span><span class="p">)</span>
<span class="k">def</span> <span class="nf">process</span><span class="p">(</span><span class="n">settings</span><span class="p">,</span> <span class="n">filename</span><span class="p">):</span>
<span class="n">os</span><span class="o">.</span><span class="n">system</span><span class="p">(</span><span class="s1">&#39;shuf </span><span class="si">%s</span><span class="s1"> &gt; </span><span class="si">%s</span><span class="s1">.shuf&#39;</span> <span class="o">%</span> <span class="p">(</span><span class="n">filename</span><span class="p">,</span> <span class="n">filename</span><span class="p">))</span> <span class="c1"># shuffle before.</span>
<span class="k">with</span> <span class="nb">open</span><span class="p">(</span><span class="s1">&#39;</span><span class="si">%s</span><span class="s1">.shuf&#39;</span> <span class="o">%</span> <span class="n">filename</span><span class="p">,</span> <span class="s1">&#39;r&#39;</span><span class="p">)</span> <span class="k">as</span> <span class="n">f</span><span class="p">:</span>
<span class="k">for</span> <span class="n">line</span> <span class="ow">in</span> <span class="n">f</span><span class="p">:</span>
<span class="k">yield</span> <span class="n">get_sample_from_line</span><span class="p">(</span><span class="n">line</span><span class="p">)</span>
</pre></div>
</div>
<p>同时 <code class="code docutils literal"><span class="pre">&#64;provider</span></code> 接口有一个 <code class="code docutils literal"><span class="pre">cache</span></code> 参数来控制缓存方法，将其设置成 <code class="code docutils literal"><span class="pre">CacheType.CACHE_PASS_IN_MEM</span></code> 的话，会将第一个 <code class="code docutils literal"><span class="pre">pass</span></code> (过完所有训练数据即为一个pass)生成的数据缓存在内存里，在之后的 <code class="code docutils literal"><span class="pre">pass</span></code> 中，不会再从 <code class="code docutils literal"><span class="pre">python</span></code> 端读取数据，而是直接从内存的缓存里读取数据。这也会极大减少数据读入的耗时。</p>
</div>
<div class="section" id="id7">
<h3><a class="toc-backref" href="#id20">加速训练速度</a><a class="headerlink" href="#id7" title="Permalink to this headline"></a></h3>
<p>PaddlePaddle支持Sparse的训练，sparse训练需要训练特征是 <code class="code docutils literal"><span class="pre">sparse_binary_vector</span></code> 、 <code class="code docutils literal"><span class="pre">sparse_vector</span></code> 、或者 <code class="code docutils literal"><span class="pre">integer_value</span></code> 的任一一种。同时，与这个训练数据交互的Layer，需要将其Parameter设置成 sparse 更新模式，即设置 <code class="code docutils literal"><span class="pre">sparse_update=True</span></code></p>
<p>这里使用简单的 <code class="code docutils literal"><span class="pre">word2vec</span></code> 训练语言模型距离，具体使用方法为:</p>
<p>使用一个词前两个词和后两个词，来预测这个中间的词。这个任务的DataProvider为:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">DICT_DIM</span><span class="o">=</span><span class="mi">3000</span>
<span class="nd">@provider</span><span class="p">(</span><span class="n">input_types</span><span class="o">=</span><span class="p">[</span><span class="n">integer_sequence</span><span class="p">(</span><span class="n">DICT_DIM</span><span class="p">),</span> <span class="n">integer_value</span><span class="p">(</span><span class="n">DICT_DIM</span><span class="p">)])</span>
<span class="k">def</span> <span class="nf">process</span><span class="p">(</span><span class="n">settings</span><span class="p">,</span> <span class="n">filename</span><span class="p">):</span>
<span class="k">with</span> <span class="nb">open</span><span class="p">(</span><span class="n">filename</span><span class="p">)</span> <span class="k">as</span> <span class="n">f</span><span class="p">:</span>
<span class="c1"># yield word ids to predict inner word id</span>
<span class="c1"># such as [28, 29, 10, 4], 4</span>
<span class="c1"># It means the sentance is 28, 29, 4, 10, 4.</span>
<span class="k">yield</span> <span class="n">read_next_from_file</span><span class="p">(</span><span class="n">f</span><span class="p">)</span>
</pre></div>
</div>
<p>这个任务的配置为:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="o">...</span> <span class="c1"># the settings and define data provider is omitted.</span>
<span class="n">DICT_DIM</span><span class="o">=</span><span class="mi">3000</span> <span class="c1"># dictionary dimension.</span>
<span class="n">word_ids</span><span class="o">=</span><span class="n">data_layer</span><span class="p">(</span><span class="s1">&#39;word_ids&#39;</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="n">DICT_DIM</span><span class="p">)</span>
<span class="n">emb</span> <span class="o">=</span> <span class="n">embedding_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">word_ids</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">256</span><span class="p">,</span> <span class="n">param_attr</span><span class="o">=</span><span class="n">ParamAttr</span><span class="p">(</span><span class="n">sparse_update</span><span class="o">=</span><span class="kc">True</span><span class="p">))</span>
<span class="n">emb_sum</span> <span class="o">=</span> <span class="n">pooling_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">emb</span><span class="p">,</span> <span class="n">pooling_type</span><span class="o">=</span><span class="n">SumPooling</span><span class="p">())</span>
<span class="n">predict</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">emb_sum</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="n">DICT_DIM</span><span class="p">,</span> <span class="n">act</span><span class="o">=</span><span class="n">Softmax</span><span class="p">())</span>
<span class="n">outputs</span><span class="p">(</span><span class="n">classification_cost</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">predict</span><span class="p">,</span> <span class="n">label</span><span class="o">=</span><span class="n">data_layer</span><span class="p">(</span><span class="s1">&#39;label&#39;</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="n">DICT_DIM</span><span class="p">)))</span>
</pre></div>
</div>
<p>更多关于sparse训练的内容请参考 <a class="reference external" href="TBD">sparse训练的文档</a></p>
</div>
<div class="section" id="id8">
<h3><a class="toc-backref" href="#id21">利用更多的计算资源</a><a class="headerlink" href="#id8" title="Permalink to this headline"></a></h3>
<p>利用更多的计算资源可以分为一下几个方式来进行:</p>
<ul class="simple">
<li>单机CPU训练
* 使用多线程训练。设置命令行参数 <code class="code docutils literal"><span class="pre">trainer_count</span></code>，即可以设置参与训练的线程数量。使用方法为 <code class="code docutils literal"><span class="pre">paddle</span> <span class="pre">train</span> <span class="pre">--trainer_count=4</span></code></li>
<li>单机GPU训练
* 使用显卡训练。设置命令行参数 <code class="code docutils literal"><span class="pre">use_gpu</span></code>。 使用方法为 <code class="code docutils literal"><span class="pre">paddle</span> <span class="pre">train</span> <span class="pre">--use_gpu=true</span></code>
* 使用多块显卡训练。设置命令行参数 <code class="code docutils literal"><span class="pre">use_gpu</span></code> 和 <code class="code docutils literal"><span class="pre">trainer_count</span></code>。使用 <code class="code docutils literal"><span class="pre">--use_gpu=True</span></code> 开启GPU训练，使用 <code class="code docutils literal"><span class="pre">trainer_count</span></code> 指定显卡数量。使用方法为 <code class="code docutils literal"><span class="pre">paddle</span> <span class="pre">train</span> <span class="pre">--use_gpu=true</span> <span class="pre">--trainer_count=4</span></code></li>
<li>多机训练
* 使用多机训练的方法也比较简单，需要先在每个节点启动 <code class="code docutils literal"><span class="pre">paddle</span> <span class="pre">pserver</span></code>，在使用 <code class="code docutils literal"><span class="pre">paddle</span> <span class="pre">train</span> <span class="pre">--pservers=192.168.100.1,192.168.100.2</span></code> 来指定每个pserver的ip地址
* 具体的多机训练方法参考 <a class="reference external" href="TBD">多机训练</a> 文档。</li>
</ul>
</div>
</div>
<div class="section" id="illegal-instruction">
<h2><a class="toc-backref" href="#id22">3. 遇到“非法指令”或者是“illegal instruction”</a><a class="headerlink" href="#illegal-instruction" title="Permalink to this headline"></a></h2>
<p>paddle在进行计算的时候为了提升计算性能，使用了avx指令。部分老的cpu型号无法支持这样的指令。通常来说执行下grep avx /proc/cpuinfo看看是否有输出即可知道是否支持。（另：用此方法部分虚拟机可能检测到支持avx指令但是实际运行会挂掉，请当成是不支持，看下面的解决方案）</p>
<p>解决办法是:</p>
<ul class="simple">
<li>使用 NO_AVX的 <a class="reference external" href="../build_and_install/index.html">安装包</a> 或者 <a class="reference external" href="../build_and_install/install/docker_install.html">Docker image</a></li>
<li>或者，使用 <code class="code docutils literal"><span class="pre">-DWITH_AVX=OFF</span></code> 重新编译PaddlePaddle。</li>
</ul>
</div>
<div class="section" id="sgd">
<h2><a class="toc-backref" href="#id23">4. 如何选择SGD算法的学习率</a><a class="headerlink" href="#sgd" title="Permalink to this headline"></a></h2>
<p>在采用sgd/async_sgd进行训练时，一个重要的问题是选择正确的learning_rate。如果learning_rate太大，那么训练有可能不收敛，如果learning_rate太小，那么收敛可能很慢，导致训练时间过长。</p>
<p>通常做法是从一个比较大的learning_rate开始试，如果不收敛，那减少学习率10倍继续试验，直到训练收敛为止。那么如何判断训练不收敛呢？可以估计出如果模型采用不变的输出最小的cost0是多少。</p>
<p>如果训练过程的的cost明显高于这个常数输出的cost，那么我们可以判断为训练不收敛。举一个例子，假如我们是三分类问题，采用multi-class-cross-entropy作为cost，数据中0,1,2三类的比例为 <code class="code docutils literal"><span class="pre">0.2,</span> <span class="pre">0.5,</span> <span class="pre">0.3</span></code> , 那么常数输出所能达到的最小cost是 <code class="code docutils literal"><span class="pre">-(0.2*log(0.2)+0.5*log(0.5)+0.3*log(0.3))=1.03</span></code> 。如果训练一个pass（或者更早）后，cost还大于这个数，那么可以认为训练不收敛，应该降低学习率。</p>
</div>
<div class="section" id="id11">
<h2><a class="toc-backref" href="#id24">5. 如何初始化参数</a><a class="headerlink" href="#id11" title="Permalink to this headline"></a></h2>
<p>默认情况下，PaddlePaddle使用均值0，标准差为 <span class="math">\(\frac{1}{\sqrt{d}}\)</span> 来初始化参数。其中 <span class="math">\(d\)</span> 为参数矩阵的宽度。这种初始化方式在一般情况下不会产生很差的结果。如果用户想要自定义初始化方式，PaddlePaddle目前提供两种参数初始化的方式:</p>
<ul class="simple">
<li>高斯分布。将 <code class="code docutils literal"><span class="pre">param_attr</span></code> 设置成 <code class="code docutils literal"><span class="pre">param_attr=ParamAttr(initial_mean=0.0,</span> <span class="pre">initial_std=1.0)</span></code></li>
<li>均匀分布。将 <code class="code docutils literal"><span class="pre">param_attr</span></code> 设置成 <code class="code docutils literal"><span class="pre">param_attr=ParamAttr(initial_max=1.0,</span> <span class="pre">initial_min=-1.0)</span></code></li>
</ul>
<p>比如设置一个全连接层的参数初始化方式和bias初始化方式，可以使用如下代码。</p>
<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">hidden</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">ipt</span><span class="p">,</span> <span class="n">param_attr</span><span class="o">=</span><span class="n">ParamAttr</span><span class="p">(</span><span class="n">initial_max</span><span class="o">=</span><span class="mf">1.0</span><span class="p">,</span> <span class="n">initial_min</span><span class="o">=-</span><span class="mf">1.0</span><span class="p">),</span>
<span class="n">bias_attr</span><span class="o">=</span><span class="n">ParamAttr</span><span class="p">(</span><span class="n">initial_mean</span><span class="o">=</span><span class="mf">1.0</span><span class="p">,</span> <span class="n">initial_std</span><span class="o">=</span><span class="mf">0.0</span><span class="p">))</span>
</pre></div>
</div>
<p>上述代码将bias全部初始化为1.0, 同时将参数初始化为 <code class="code docutils literal"><span class="pre">[1.0,</span> <span class="pre">-1.0]</span></code> 的均匀分布。</p>
</div>
<div class="section" id="id12">
<h2><a class="toc-backref" href="#id25">6. 如何共享参数</a><a class="headerlink" href="#id12" title="Permalink to this headline"></a></h2>
<p>PaddlePaddle的参数使用名字 <code class="code docutils literal"><span class="pre">name</span></code> 作为参数的ID，相同名字的参数，会共享参数。设置参数的名字，可以使用 <code class="code docutils literal"><span class="pre">ParamAttr(name=&quot;YOUR_PARAM_NAME&quot;)</span></code> 来设置。更方便的设置方式，是想要共享的参数使用同样的 <code class="code docutils literal"><span class="pre">ParamAttr</span></code> 对象。</p>
<p>简单的全连接网络，参数共享的配置示例为:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="kn">from</span> <span class="nn">paddle.trainer_config_helpers</span> <span class="k">import</span> <span class="o">*</span>
<span class="n">settings</span><span class="p">(</span>
<span class="n">learning_rate</span><span class="o">=</span><span class="mi">1</span><span class="n">e</span><span class="o">-</span><span class="mi">4</span><span class="p">,</span>
<span class="n">batch_size</span><span class="o">=</span><span class="mi">1000</span>
<span class="p">)</span>
<span class="n">a</span> <span class="o">=</span> <span class="n">data_layer</span><span class="p">(</span><span class="n">name</span><span class="o">=</span><span class="s1">&#39;feature_a&#39;</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">200</span><span class="p">)</span>
<span class="n">b</span> <span class="o">=</span> <span class="n">data_layer</span><span class="p">(</span><span class="n">name</span><span class="o">=</span><span class="s1">&#39;feature_b&#39;</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">200</span><span class="p">)</span>
<span class="n">fc_param</span> <span class="o">=</span> <span class="n">ParamAttr</span><span class="p">(</span><span class="n">name</span><span class="o">=</span><span class="s1">&#39;fc_param&#39;</span><span class="p">,</span> <span class="n">initial_max</span><span class="o">=</span><span class="mf">1.0</span><span class="p">,</span> <span class="n">initial_min</span><span class="o">=-</span><span class="mf">1.0</span><span class="p">)</span>
<span class="n">bias_param</span> <span class="o">=</span> <span class="n">ParamAttr</span><span class="p">(</span><span class="n">name</span><span class="o">=</span><span class="s1">&#39;bias_param&#39;</span><span class="p">,</span> <span class="n">initial_mean</span><span class="o">=</span><span class="mf">0.0</span><span class="p">,</span> <span class="n">initial_std</span><span class="o">=</span><span class="mf">0.0</span><span class="p">)</span>
<span class="n">softmax_param</span> <span class="o">=</span> <span class="n">ParamAttr</span><span class="p">(</span><span class="n">name</span><span class="o">=</span><span class="s1">&#39;softmax_param&#39;</span><span class="p">,</span> <span class="n">initial_max</span><span class="o">=</span><span class="mf">1.0</span><span class="p">,</span> <span class="n">initial_min</span><span class="o">=-</span><span class="mf">1.0</span><span class="p">)</span>
<span class="n">hidden_a</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">a</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">200</span><span class="p">,</span> <span class="n">param_attr</span><span class="o">=</span><span class="n">fc_param</span><span class="p">,</span> <span class="n">bias_attr</span><span class="o">=</span><span class="n">bias_param</span><span class="p">)</span>
<span class="n">hidden_b</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">b</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">200</span><span class="p">,</span> <span class="n">param_attr</span><span class="o">=</span><span class="n">fc_param</span><span class="p">,</span> <span class="n">bias_attr</span><span class="o">=</span><span class="n">bias_param</span><span class="p">)</span>
<span class="n">predict</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="p">[</span><span class="n">hidden_a</span><span class="p">,</span> <span class="n">hidden_b</span><span class="p">],</span> <span class="n">param_attr</span><span class="o">=</span><span class="p">[</span><span class="n">softmax_param</span><span class="p">,</span> <span class="n">softmax_param</span><span class="p">],</span>
<span class="n">bias_attr</span><span class="o">=</span><span class="kc">False</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">10</span><span class="p">,</span> <span class="n">act</span><span class="o">=</span><span class="n">SoftmaxActivation</span><span class="p">())</span>
<span class="n">outputs</span><span class="p">(</span><span class="n">classification_cost</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">predict</span><span class="p">,</span> <span class="n">label</span><span class="o">=</span><span class="n">data_layer</span><span class="p">(</span><span class="n">name</span><span class="o">=</span><span class="s1">&#39;label&#39;</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">10</span><span class="p">)))</span>
</pre></div>
</div>
<p>这里 <code class="code docutils literal"><span class="pre">hidden_a</span></code> 和 <code class="code docutils literal"><span class="pre">hidden_b</span></code> 使用了同样的parameter和bias。并且softmax层的两个输入也使用了同样的参数 <code class="code docutils literal"><span class="pre">softmax_param</span></code>。</p>
</div>
</div>
</div>
</div>
</div>
<div class="sphinxsidebar" role="navigation" aria-label="main navigation">
<div class="sphinxsidebarwrapper">
<h3><a href="../index.html">Table Of Contents</a></h3>
<ul>
<li><a class="reference internal" href="#">PaddlePaddle常见问题</a><ul>
<li><a class="reference internal" href="#id1">1. 如何减少PaddlePaddle的内存占用</a><ul>
<li><a class="reference internal" href="#dataprovider">减少DataProvider缓冲池内存</a></li>
<li><a class="reference internal" href="#id3">神经元激活内存</a></li>
<li><a class="reference internal" href="#id4">参数内存</a></li>
</ul>
</li>
<li><a class="reference internal" href="#id5">2. 如何加速PaddlePaddle的训练速度</a><ul>
<li><a class="reference internal" href="#id6">减少数据载入的耗时</a></li>
<li><a class="reference internal" href="#id7">加速训练速度</a></li>
<li><a class="reference internal" href="#id8">利用更多的计算资源</a></li>
</ul>
</li>
<li><a class="reference internal" href="#illegal-instruction">3. 遇到“非法指令”或者是“illegal instruction”</a></li>
<li><a class="reference internal" href="#sgd">4. 如何选择SGD算法的学习率</a></li>
<li><a class="reference internal" href="#id11">5. 如何初始化参数</a></li>
<li><a class="reference internal" href="#id12">6. 如何共享参数</a></li>
</ul>
</li>
</ul>
<div role="note" aria-label="source link">
<h3>This Page</h3>
<ul class="this-page-menu">
<li><a href="../_sources/faq/index.txt"
rel="nofollow">Show Source</a></li>
</ul>
</div>
<div id="searchbox" style="display: none" role="search">
<h3>Quick search</h3>
<form class="search" action="../search.html" method="get">
<div><input type="text" name="q" /></div>
<div><input type="submit" value="Go" /></div>
<input type="hidden" name="check_keywords" value="yes" />
<input type="hidden" name="area" value="default" />
</form>
</div>
<script type="text/javascript">$('#searchbox').show(0);</script>
</div>
</div>
<div class="clearer"></div>
</div>
<div class="related" role="navigation" aria-label="related navigation">
<h3>Navigation</h3>
<ul>
<li class="right" style="margin-right: 10px">
<a href="../genindex.html" title="General Index"
>index</a></li>
<li class="nav-item nav-item-0"><a href="../index.html">PaddlePaddle documentation</a> &#187;</li>
</ul>
</div>
<div class="footer" role="contentinfo">
&#169; Copyright 2016, PaddlePaddle developers.
Created using <a href="http://sphinx-doc.org/">Sphinx</a> 1.4.8.
</div>
</body>
</html>
\ No newline at end of file
doc_cn/index.html
浏览文件 @ 8f5e20e7
......@@ -84,12 +84,6 @@ var _hmt = _hmt || [];
<li><a class="reference external" href="algorithm/rnn/hierarchical-layer.html">支持双层序列作为输入的Layer</a></li>
</ul>
</div>
<div class="section" id="id12">
<h2>常见问题<a class="headerlink" href="#id12" title="Permalink to this headline"></a></h2>
<ul class="simple">
<li><a class="reference external" href="faq/index.html">常见问题</a></li>
</ul>
</div>
</div>
......@@ -104,7 +98,6 @@ var _hmt = _hmt || [];
<li><a class="reference internal" href="#id1">使用指南</a></li>
<li><a class="reference internal" href="#id8">开发指南</a></li>
<li><a class="reference internal" href="#id9">算法教程</a></li>
<li><a class="reference internal" href="#id12">常见问题</a></li>
</ul>
</li>
</ul>
......
doc_cn/objects.inv
浏览文件 @ 8f5e20e7
无法预览此类型文件
doc_cn/searchindex.js
浏览文件 @ 8f5e20e7
Search.setIndex({envversion:49,filenames:["algorithm/rnn/hierarchical-layer","algorithm/rnn/hierarchical-rnn","algorithm/rnn/rnn-tutorial","build/docker/build_docker_image","build_and_install/cmake/compile_options","build_and_install/cmake/index","build_and_install/cmake/install_deps","build_and_install/cmake/make_and_install","build_and_install/index","build_and_install/install/docker_install","build_and_install/install/ubuntu_install","cluster/index","demo/index","demo/quick_start/index","faq/index","index","ui/cmd/dump_config","ui/cmd/index","ui/cmd/make_diagram","ui/cmd/merge_model","ui/cmd/paddle_pserver","ui/cmd/paddle_train","ui/cmd/paddle_version","ui/data_provider/index","ui/data_provider/pydataprovider2","ui/data_provider/write_new_dataprovider","ui/index","ui/predict/swig_py_paddle"],objects:{},objnames:{},objtypes:{},terms:{"0000x":13,"000\u5f20\u7070\u5ea6\u56fe\u7247\u7684\u6570\u5b57\u5206\u7c7b\u6570\u636e\u96c6":24,"00186201e":27,"04\u4e2d\u6b63\u786e":10,"08823112e":27,"0\u5c42\u5e8f\u5217":0,"0b1":10,"100m":14,"10\u4ee5\u4e0a\u7684linux":9,"10\u7ef4\u7684\u6574\u6570\u503c":24,"10gbe":9,"10m":3,"12194102e":27,"12\u7248\u672c\u6d4b\u8bd5\u901a\u8fc7":3,"12\u7248\u672c\u7684\u60c5\u51b5\u4e0b\u5e76\u6ca1\u6709\u6d4b\u8bd5":3,"15501715e":27,"15mb":13,"16mb":13,"1\u7684\u8bdd":24,"252kb":13,"25639710e":27,"25k":13,"27787406e":27,"28\u7684\u50cf\u7d20\u7070\u5ea6\u503c":24,"28\u7684\u7a20\u5bc6\u5411\u91cf\u548c\u4e00\u4e2a":24,"2\u4e09\u7c7b\u7684\u6bd4\u4f8b\u4e3a":14,"2\u4e2a\u5b50\u53e5":1,"2\u53e5\u53cc\u5c42\u5e8f\u5217\u548c5\u53e5\u5355\u5c42\u5e8f\u5217\u7684\u6570\u636e\u5b8c\u5168\u4e00\u6837":1,"2\u8868\u793a\u4e00\u6b21\u8fc72\u53e5\u53cc\u5c42\u5e8f\u5217":1,"2\u8fdb\u884c\u8fdb\u4e00\u6b65\u6f14\u5316":13,"32777140e":27,"36540484e":27,"3\u4e2a\u5b50\u53e5":1,"40gbe":9,"43630644e":27,"48565123e":27,"48684503e":27,"49316648e":27,"500m":14,"50k":3,"51111044e":27,"53018653e":27,"56gbe":9,"5\u5230\u672c\u5730\u73af\u5883\u4e2d":10,"5\u8868\u793a\u4e00\u6b21\u8fc75\u53e5\u5355\u5c42\u5e8f\u5217":1,"70634608e":27,"72296313e":27,"85625684e":27,"93137714e":27,"96644767e":27,"99982715e":27,"9\u7684\u6570\u5b57":24,"\u4e00":1,"\u4e00\u4e2a0\u5c42\u5e8f\u5217":0,"\u4e00\u4e2a0\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u6269\u5c55\u6210\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u6269\u5c55\u6210\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":0,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u8fdb\u5165":2,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":0,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u6216\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":0,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u8fdb\u5165":2,"\u4e00\u4e2a\u53cc\u5c42rnn\u7531\u591a\u4e2a\u5355\u5c42rnn\u7ec4\u6210":2,"\u4e00\u4e2a\u53ef\u8c03\u7528\u7684\u51fd\u6570":2,"\u4e00\u4e2a\u6570\u636e\u96c6\u5927\u90e8\u5206\u5e8f\u5217\u957f\u5ea6\u662f100":14,"\u4e00\u4e2a\u6587\u4ef6":24,"\u4e00\u4e2a\u72ec\u7acb\u7684\u5143\u7d20":0,"\u4e00\u4e2a\u72ec\u7acb\u7684\u8bcd\u8bed":0,"\u4e00\u4e2a\u91cd\u8981\u7684\u95ee\u9898\u662f\u9009\u62e9\u6b63\u786e\u7684learning_r":14,"\u4e00\u4e2alabel":1,"\u4e00\u4e2alogging\u5bf9\u8c61":24,"\u4e00\u4e2apass\u8868\u793a\u8fc7\u4e00\u904d\u6240\u6709\u8bad\u7ec3\u6837\u672c":13,"\u4e00\u4eba":1,"\u4e00\u5171\u670910\u4e2a\u6837\u672c":1,"\u4e00\u5171\u67094\u4e2a\u6837\u672c":1,"\u4e00\u53e5\u8bdd\u662f\u7531\u8bcd\u8bed\u6784\u6210\u7684\u5e8f\u5217":2,"\u4e00\u65e9":1,"\u4e00\u662fbatch":14,"\u4e00\u6761":24,"\u4e00\u6b21\u6027\u676f\u5b50":1,"\u4e00\u81f4":1,"\u4e00\u81f4\u7684\u7279\u5f81":24,"\u4e00\u822c\u60c5\u51b5\u4e0b":23,"\u4e00\u822c\u63a8\u8350\u8bbe\u7f6e\u6210true":24,"\u4e00\u884c\u4e3a\u4e00\u4e2a\u6837\u672c":13,"\u4e00\u884c\u5bf9\u5e94\u4e00\u4e2a\u6570\u636e\u6587\u4ef6":23,"\u4e0a\u7684\u6548\u679c":13,"\u4e0a\u7f51":1,"\u4e0a\u8ff0\u4ee3\u7801\u5c06bias\u5168\u90e8\u521d\u59cb\u5316\u4e3a1":14,"\u4e0b\u6587\u4ee5nlp\u4efb\u52a1\u4e3a\u4f8b":2,"\u4e0b\u6b21":1,"\u4e0b\u8f7d\u8fdb\u7a0b\u4f1a\u91cd\u542f":3,"\u4e0b\u8ff0\u5185\u5bb9\u5c06\u5206\u4e3a\u5982\u4e0b\u51e0\u4e2a\u7c7b\u522b\u63cf\u8ff0":9,"\u4e0b\u975e\u5e38\u5c11\u7684\u53d8\u91cf\u5f15\u7528":24,"\u4e0b\u9762\u8fd9\u4e9blayer\u80fd\u591f\u63a5\u53d7\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165":0,"\u4e0b\u9762dataprovid":13,"\u4e0d":1,"\u4e0d\u4e00\u5b9a\u548c\u65f6\u95f4\u6709\u5173\u7cfb":24,"\u4e0d\u4f1a\u518d\u4ece":14,"\u4e0d\u4f1a\u6267\u884c\u6d4b\u8bd5\u64cd\u4f5c":23,"\u4e0d\u5305\u542blabel":27,"\u4e0d\u540c\u7684\u4f18\u5316\u7b97\u6cd5\u9700\u8981\u4f7f\u7528\u4e0d\u540c\u5927\u5c0f\u7684\u5185\u5b58":14,"\u4e0d\u540c\u7684\u6570\u636e\u7c7b\u578b\u548c\u5e8f\u5217\u6a21\u5f0f\u8fd4\u56de\u7684\u683c\u5f0f\u4e0d\u540c":24,"\u4e0d\u540c\u8f93\u5165\u542b\u6709\u7684\u5b50\u53e5":2,"\u4e0d\u540c\u8f93\u5165\u5e8f\u5217\u542b\u6709\u7684\u8bcd\u8bed\u6570\u5fc5\u987b\u4e25\u683c\u76f8\u7b49":2,"\u4e0d\u5c11":1,"\u4e0d\u5e94\u8be5\u88ab\u62c6\u89e3":2,"\u4e0d\u6307\u5b9a\u65f6":2,"\u4e0d\u652f\u6301avx\u6307\u4ee4\u96c6\u7684cpu\u4e5f\u53ef\u4ee5\u8fd0\u884c":9,"\u4e0d\u7f13\u5b58\u4efb\u4f55\u6570\u636e":24,"\u4e0d\u8fc7":1,"\u4e0d\u8fdc":1,"\u4e0d\u9519":1,"\u4e0d\u9700\u8981avx\u6307\u4ee4\u96c6\u7684cpu\u4e5f\u53ef\u4ee5\u8fd0\u884c":9,"\u4e0e\u8bad\u7ec3\u7f51\u7edc\u914d\u7f6e\u4e0d\u540c\u7684\u662f":13,"\u4e0e\u8fd9\u4e2a\u8bad\u7ec3\u6570\u636e\u4ea4\u4e92\u7684layer":14,"\u4e14":1,"\u4e14\u5e8f\u5217\u7684\u6bcf\u4e00\u4e2a\u5143\u7d20\u8fd8\u662f\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":24,"\u4e24":1,"\u4e24\u4e2a\u5d4c\u5957\u7684":2,"\u4e24\u4e2a\u6587\u6863":9,"\u4e24\u7c7b":13,"\u4e25\u91cd\u7684\u95ee\u9898":24,"\u4e2a":13,"\u4e2a\u5185\u5b58\u6c60\u5b9e\u9645\u4e0a\u51b3\u5b9a\u4e86shuffle\u7684\u7c92\u5ea6":14,"\u4e2ayield":24,"\u4e2d":[13,14],"\u4e2d\u5b9a\u4e49\u4f7f\u7528\u54ea\u79cddataprovider\u53ca\u5176\u53c2\u6570":23,"\u4e2d\u5b9a\u4e49\u548c\u4f7f\u7528":2,"\u4e2d\u5b9a\u4e49\u7684\u987a\u5e8f\u4e00\u81f4":24,"\u4e2d\u5bfb\u627e\u8fd9\u4e9bblas\u7684\u5b9e\u73b0":4,"\u4e2d\u7684":24,"\u4e2d\u7684\u4e8c\u8fdb\u5236\u4f7f\u7528\u4e86":9,"\u4e2d\u7684set":24,"\u4e2d\u914d\u7f6e":24,"\u4e34\u65f6\u53d8\u91cf\u7b49\u7b49":14,"\u4e3a":24,"\u4e3a\u4e86\u63cf\u8ff0\u65b9\u4fbf":2,"\u4e3a\u4e86\u8fd0\u884cpaddlepaddle\u7684docker\u955c\u50cf":9,"\u4e3a\u4f8b\u8fdb\u884c\u9884\u6d4b":13,"\u4e3a\u53c2\u6570\u77e9\u9635\u7684\u5bbd\u5ea6":14,"\u4e3b\u8981\u51fd\u6570\u662fprocess\u51fd\u6570":24,"\u4e3b\u8981\u5206\u4e3a\u4ee5\u4e0b\u51e0\u4e2a\u6b65\u9aa4":27,"\u4e3b\u8981\u5305\u62ec\u4e24\u90e8\u5206":13,"\u4e3b\u8981\u539f\u56e0":1,"\u4e3b\u8981\u662f\u589e\u52a0\u4e86\u521d\u59cb\u5316\u673a\u5236":24,"\u4e3b\u8981\u6b65\u9aa4\u4e3a":27,"\u4e3b\u8981\u7531\u4e8e\u65e7\u7248\u672c":3,"\u4e3b\u8981\u7684\u8f6f\u4ef6\u5305\u4e3apy_paddl":27,"\u4e3e\u4e00\u4e2a\u4f8b\u5b50":14,"\u4e4b\u95f4\u7684\u8fd0\u7b97\u662f\u72ec\u7acb\u7684":2,"\u4e5f":1,"\u4e5f\u4e0d\u5b58\u5728\u4e00\u4e2asubseq\u76f4\u63a5\u751f\u6210\u4e0b\u4e00\u4e2asubseq\u7684\u60c5\u51b5":2,"\u4e5f\u4f1a\u6254\u5230\u8fd9\u6761\u6570\u636e":24,"\u4e5f\u4f1a\u8bfb\u53d6\u76f8\u5173\u8def\u5f84\u53d8\u91cf\u6765\u8fdb\u884c\u641c\u7d22":4,"\u4e5f\u53ef\u4ee5":24,"\u4e5f\u53ef\u4ee5\u4e3ainteg":1,"\u4e5f\u53ef\u4ee5\u4f7f\u7528":24,"\u4e5f\u53ef\u4ee5\u548cpaddl":17,"\u4e5f\u53ef\u4ee5\u662f\u4e00\u4e2a\u8bcd\u8bed":2,"\u4e5f\u53ef\u4ee5\u76f4\u63a5\u6267\u884c":9,"\u4e5f\u53ef\u4ee5\u901a\u8fc7\u5982\u4e0b\u65b9\u5f0f\u9884\u6d4b":13,"\u4e5f\u53ef\u4ee5\u901a\u8fc7save":13,"\u4e5f\u53ef\u4ee5\u9884\u6d4b\u6ca1\u6709label\u7684\u6d4b\u8bd5\u96c6":13,"\u4e5f\u5c31\u662f\u5c06\u8bcd\u5411\u91cf\u6a21\u578b\u989d\u6b65":13,"\u4e5f\u5c31\u662f\u76f4\u63a5\u5199\u5185\u5b58\u7684float\u6570\u7ec4":27,"\u4e5f\u662fdecoder\u5faa\u73af\u5c55\u5f00\u7684\u4f9d\u636e":2,"\u4e5f\u9700\u8981\u4e24\u6b21\u968f\u673a\u9009\u62e9\u5230\u540c\u6837\u7684generator\u7684\u65f6\u5019":24,"\u4e7e":1,"\u4e86":1,"\u4e86\u975e\u5e38\u65b9\u4fbf\u7684\u4e8c\u8fdb\u5236\u5206\u53d1\u624b\u6bb5":9,"\u4e8c\u6b21\u5f00\u53d1\u53ef\u4ee5":9,"\u4e94\u661f\u7ea7":1,"\u4ea4\u901a":1,"\u4ea4\u901a\u4fbf\u5229":1,"\u4eba\u5458\u7b49\u7b49":3,"\u4ec5\u4ec5\u9700\u8981":24,"\u4ecb\u7ecdpaddlepaddle\u4f7f\u7528\u6d41\u7a0b\u548c\u5e38\u7528\u7684\u7f51\u7edc\u57fa\u7840\u5355\u5143\u7684\u914d\u7f6e\u65b9\u6cd5":13,"\u4ece\u4e00\u4e2aword\u751f\u6210\u4e0b\u4e00\u4e2aword":2,"\u4ece\u6587\u4ef6\u4e2d\u8bfb\u53d6\u6bcf\u4e00\u6761\u6570\u636e":24,"\u4ece\u6587\u672c\u6587\u4ef6\u4e2d\u8bfb\u53d6":24,"\u4ece\u800c\u4e0d\u80fd\u5728\u8fd0\u884c\u7f16\u8bd1\u547d\u4ee4\u7684\u65f6\u5019\u63a5\u53d7\u53c2\u6570":3,"\u4ece\u800c\u751f\u6210\u591a\u4e2agener":24,"\u4ece\u800c\u9632\u6b62\u8fc7\u62df\u5408":23,"\u4ece\u8bed\u4e49\u4e0a\u770b":2,"\u4ece\u8f93\u5165\u6570\u636e\u4e0a\u770b":1,"\u4ed6\u4eec\u662f":[9,10,14,17,24],"\u4ed6\u4eec\u7684imag":9,"\u4ed6\u53ef\u4ee5\u5c06\u67d0\u4e00\u4e2a\u51fd\u6570\u6807\u8bb0\u6210\u4e00\u4e2apydataprovid":24,"\u4ee3\u8868\u4e00\u4e2a\u5411\u91cf":1,"\u4ee3\u8868\u4e0d\u540c\u7684\u53cc\u5c42\u5e8f\u5217":1,"\u4ee3\u8868\u5355\u5c42\u5e8f\u5217":1,"\u4ee3\u8868\u53cc\u5c42\u5e8f\u5217":1,"\u4ee4\u884c\u53c2\u6570\u6587\u6863":13,"\u4ee5\u53ca\u53cc\u5c42\u5e8f\u5217":0,"\u4ee5\u53ca\u8ba1\u7b97\u903b\u8f91\u5728\u5e8f\u5217\u4e0a\u7684\u5faa\u73af\u5c55\u5f00":2,"\u4ee5\u592a\u7f51\u5361":9,"\u4ee5\u76f8\u5bf9\u8def\u5f84\u5f15\u7528":23,"\u4ef7\u683c":1,"\u4efb\u610f\u4e00\u79cdcblas\u5b9e\u73b0":4,"\u4f1a\u5171\u4eab\u53c2\u6570":14,"\u4f1a\u5bf9\u6bcf\u4e00\u4e2a\u6fc0\u6d3b\u6682\u5b58\u4e00\u4e9b\u6570\u636e":14,"\u4f1a\u5bf9\u8fd9\u7c7b\u8f93\u5165\u8fdb\u884c\u62c6\u89e3":2,"\u4f1a\u5c06\u6bcf\u4e2a\u65f6\u95f4\u6b65\u7684\u8f93\u51fa\u62fc\u63a5":2,"\u4f1a\u5c06\u7b2c\u4e00\u4e2a":14,"\u4f1a\u6210\u4e3astep\u51fd\u6570\u7684\u8f93\u5165":2,"\u4f1a\u62a5\u5bfb\u627e\u4e0d\u5230\u8fd9\u4e9b\u52a8\u6001\u5e93":10,"\u4f1a\u62a5\u9519":2,"\u4f1a\u6839\u636e\u547d\u4ee4\u884c\u53c2\u6570\u6307\u5b9a\u7684\u6d4b\u8bd5\u65b9\u5f0f":23,"\u4f1a\u6839\u636einput_types\u68c0\u67e5\u6570\u636e\u7684\u5408\u6cd5\u6027":24,"\u4f1a\u751f\u6210\u591a\u4e2agener":24,"\u4f1a\u88ab\u62c6\u89e3\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":2,"\u4f1a\u88ab\u62c6\u89e3\u4e3a\u975e\u5e8f\u5217":2,"\u4f1a\u9884\u5148\u8bfb\u53d6\u5168\u90e8\u6570\u636e\u5230\u5185\u5b58\u4e2d":24,"\u4f20\u5165\u547d\u4ee4\u884c\u53c2\u6570\u521d\u59cb\u5316":27,"\u4f20\u5165\u7684\u662f\u4e00\u4e2a\u51fd\u6570":24,"\u4f20\u5165\u7684\u914d\u7f6e\u53c2\u6570\u5305\u62ec":3,"\u4f20\u5165\u8fd9\u4e2a\u53d8\u91cf\u7684\u65b9\u5f0f\u4e3a":24,"\u4f46\u4ece\u4e0a\u9762\u7684\u6570\u636e\u683c\u5f0f\u53ef\u77e5":1,"\u4f46\u5b50\u53e5\u542b\u6709\u7684\u8bcd\u8bed\u6570\u53ef\u4ee5\u4e0d\u76f8\u7b49":2,"\u4f46\u662f":[3,14,24],"\u4f46\u662f\u5728":24,"\u4f46\u662f\u5982\u679c\u4f7f\u7528\u4e86\u9ad8\u6027\u80fd\u7684\u7f51\u5361":9,"\u4f46\u662f\u65b9\u4fbf\u8c03\u8bd5\u548cbenchmark":4,"\u4f46\u662f\u6709\u65f6\u4e3a\u4e86\u8ba1\u7b97\u5747\u8861\u6027":24,"\u4f46\u662f\u7a81\u7136\u6709\u4e00\u4e2a10000\u957f\u7684\u5e8f\u5217":14,"\u4f46\u662fbatch":14,"\u4f46\u7406\u8bba\u4e0a\u652f\u6301\u5176\u4ed6\u7684":10,"\u4f46\u8fd9\u79cd\u60c5\u51b5\u4e0b":1,"\u4f46\u9700\u8981\u6ce8\u610f\u7684\u662f\u7f16\u8bd1\u548c":4,"\u4f4d\u7f6e":1,"\u4f4e\u4e8edocker":3,"\u4f4f":1,"\u4f53\u53ef\u4ee5\u53c2\u8003":24,"\u4f5c\u4e3a\u4e0b\u4e00\u4e2a\u5b50\u53e5memory\u7684\u521d\u59cb\u72b6\u6001":1,"\u4f5c\u4e3a\u53c2\u6570\u7684id":14,"\u4f5c\u4e3a\u5f53\u524d\u65f6\u523b\u8f93\u5165":2,"\u4f5c\u4e3aboot":1,"\u4f5c\u7528":0,"\u4f7f\u5728python\u73af\u5883\u4e0b\u7684\u9884\u6d4b\u63a5\u53e3\u66f4\u52a0\u7b80\u5355":27,"\u4f7f\u7528":[2,4,9,14,27],"\u4f7f\u7528\u4e00\u4e2a\u8bcd\u524d\u4e24\u4e2a\u8bcd\u548c\u540e\u4e24\u4e2a\u8bcd":14,"\u4f7f\u7528\u4e86\u540c\u6837\u7684parameter\u548cbia":14,"\u4f7f\u7528\u4e86avx\u6307\u4ee4":14,"\u4f7f\u7528\u591a\u5757\u663e\u5361\u8bad\u7ec3":14,"\u4f7f\u7528\u591a\u673a\u8bad\u7ec3\u7684\u65b9\u6cd5\u4e5f\u6bd4\u8f83\u7b80\u5355":14,"\u4f7f\u7528\u591a\u7ebf\u7a0b\u8bad\u7ec3":14,"\u4f7f\u7528\u5982\u4e0b\u811a\u672c\u53ef\u4ee5\u786e\u5b9a\u672c\u673a\u7684cpu\u77e5\u5426\u652f\u6301":9,"\u4f7f\u7528\u65b9\u6cd5\u4e3a":14,"\u4f7f\u7528\u663e\u5361\u8bad\u7ec3":14,"\u4f7f\u7528\u7684\u547d\u4ee4\u4e5f\u662f":4,"\u4f7f\u7528\u8005\u53ea\u9700\u8981\u5173\u6ce8\u4e8e\u8bbe\u8ba1rnn\u5728\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185\u5b8c\u6210\u7684\u8ba1\u7b97":2,"\u4f7f\u7528\u8be5\u63a5\u53e3\u7528\u6237\u53ef\u4ee5\u53ea\u5173\u6ce8\u5982\u4f55":24,"\u4f7f\u7528\u8be5dockerfile\u6784\u5efa\u51fa\u955c\u50cf":9,"\u4f7f\u7528\u8fd9\u4e2a\u5173\u952e\u8bcd":24,"\u4f7f\u7528deb\u5305\u5728ubuntu\u4e0a\u5b89\u88c5paddlepaddl":8,"\u4f7f\u7528dockerfile\u6784\u5efa\u4e00\u4e2a\u5168\u65b0\u7684dock":9,"\u4f7f\u7528mnist\u624b\u5199\u8bc6\u522b\u4f5c\u4e3a\u6837\u4f8b":27,"\u4f7f\u7528ssh\u8bbf\u95eepaddlepaddle\u955c\u50cf":9,"\u4f86":1,"\u4f8b\u5982":[4,13,14,24],"\u4f8b\u5982\u5982\u679c\u4f7f\u7528":14,"\u4f8b\u5982\u6587\u4ef6\u540d\u662f":24,"\u4f8b\u5982rdma\u7f51\u5361":9,"\u4f8b\u5982sigmoid\u53d8\u6362":13,"\u4f9d\u6b21\u8fd4\u56de\u4e86\u6587\u4ef6\u4e2d\u7684\u6bcf\u6761\u6570\u636e":24,"\u4f9d\u7136\u4fdd\u6301\u6bcf\u4e2asubseq\u6700\u540e\u4e00\u4e2a\u5143\u7d20\u7684\u503c\u4e0d\u53d8":1,"\u4fbf\u5229":1,"\u4fbf\u5b9c":1,"\u4fe1\u606f":9,"\u5047\u5982\u6211\u4eec\u662f\u4e09\u5206\u7c7b\u95ee\u9898":14,"\u505a\u6cd5\u53ef\u4ee5\u6709\u4e24\u79cd":14,"\u505c\u7535":1,"\u5143\u7d20":0,"\u5143\u7d20\u4e4b\u95f4\u7684\u987a\u5e8f\u662f\u91cd\u8981\u7684\u8f93\u5165\u4fe1\u606f":0,"\u5168\u5bb6":1,"\u5173\u4e8edataprovider\u4e2dinput":1,"\u5173\u4e8eunbound":2,"\u5173\u95edcontain":9,"\u5176\u4e2d":[3,9,10,14,23,24,27],"\u5176\u4e2d\u6587\u672c\u8f93\u5165\u7c7b\u578b\u5b9a\u4e49\u4e3a\u6574\u6570\u65f6\u5e8f\u7c7b\u578binteg":13,"\u5176\u4e2d\u6bcf\u4e2a\u5143\u7d20\u662f\u53cc\u5c42\u5e8f\u5217\u4e2d\u6bcf\u4e2asubseq\u6700\u540e\u4e00\u4e2a":0,"\u5176\u4e2d\u7b2c\u4e00\u884c\u662f\u5f15\u5165paddlepaddle\u7684pydataprovider2\u5305":24,"\u5176\u4e2d\u7b2ci\u4e2asubseq\u4e2d\u7684\u6240\u6709\u5411\u91cf\u5747\u4e3a\u8f93\u5165\u7684\u5355\u5c42\u5e8f\u5217\u4e2d\u7684\u7b2ci\u4e2a\u5411\u91cf":1,"\u5176\u4ed6\u5185\u5b58\u6742\u9879":14,"\u5176\u4ed6\u5185\u5b58\u6742\u9879\u662f\u6307paddlepaddle\u672c\u8eab\u6240\u7528\u7684\u4e00\u4e9b\u5185\u5b58":14,"\u5176\u4ed6\u53c2\u6570\u5747\u4f7f\u7528kei":24,"\u5176\u4ed6\u53c2\u6570\u8bf7\u53c2\u8003":13,"\u5176\u4ed6\u53c2\u6570\u90fd\u4f7f\u7528kei":24,"\u5176\u4ed6\u7684\u5185\u5b58\u7684\u51cf\u5c11\u65b9\u6cd5\u4f9d\u6b21\u4e3a":14,"\u5176\u4f5c\u7528\u662f\u5c06\u8bad\u7ec3\u6570\u636e\u4f20\u5165\u5185\u5b58\u6216\u8005\u663e\u5b58":23,"\u5176\u5b83\u90e8\u5206\u548c\u903b\u8f91\u56de\u5f52\u7f51\u7edc\u7ed3\u6784\u4e00\u81f4":13,"\u5176\u5b83layer\u7684\u8f93\u51fa":2,"\u5176\u5b9e\u4e5f\u662f\u548c\u6bcf\u4e2amini":14,"\u5176\u6570\u636e\u4f7f\u7528":24,"\u5176\u6b21":1,"\u5176\u7b2c\u4e00\u884c\u8bf4\u660e\u4e86paddle\u7684\u7248\u672c":22,"\u5177":24,"\u5177\u4f53\u4f7f\u7528\u65b9\u6cd5\u4e3a":14,"\u5177\u4f53\u53ef\u4ee5\u8bbe\u7f6e\u6210\u4ec0\u4e48\u5176\u4ed6\u683c":24,"\u5177\u4f53\u53ef\u53c2\u8003\u6587\u6863":2,"\u5177\u4f53\u6709\u54ea\u4e9b\u683c\u5f0f":24,"\u5177\u4f53\u7684\u591a\u673a\u8bad\u7ec3\u65b9\u6cd5\u53c2\u8003":14,"\u5177\u4f53\u8bf7\u53c2\u8003\u6ce8\u610f\u4e8b\u9879\u4e2d\u7684":9,"\u5177\u4f53dataprovider\u8fd8\u5177\u6709\u4ec0\u4e48\u529f\u80fd":24,"\u5177\u6709\u4e24\u4e2a\u53c2\u6570":24,"\u5177\u6709\u548c\u5bbf\u4e3b\u673a\u76f8\u8fd1\u7684\u8fd0\u884c\u6548\u7387":9,"\u5177\u6709\u7684\u5c5e\u6027\u6709":24,"\u5178\u578b\u7684\u8f93\u51fa\u7ed3\u679c\u4e3a":27,"\u5178\u578b\u7684\u9884\u6d4b\u4ee3\u7801\u5982\u4e0b":27,"\u5185\u5b58\u4e0d\u591f\u7528\u7684\u60c5\u51b5":23,"\u5185\u5c42\u662f":1,"\u5185\u5c42inner":1,"\u518d\u6307\u5b9a":4,"\u5199\u5165train":24,"\u5199\u5728train":23,"\u51c6\u5907":1,"\u51c6\u5907\u6570\u636e":27,"\u51cf\u5c0f\u5e8f\u5217\u7684\u957f\u5ea6":14,"\u51cf\u5c0f\u8fd9\u4e2a\u5185\u5b58\u6c60\u5373\u53ef\u51cf\u5c0f\u5185\u5b58\u5360\u7528":14,"\u51cf\u5c0fbatch":14,"\u51fa\u53bb\u73a9":1,"\u51fa\u5dee":1,"\u51fa\u6765":1,"\u51fd\u6570":24,"\u51fd\u6570\u4e2d":24,"\u51fd\u6570\u4e2d\u4f7f\u7528":24,"\u51fd\u6570\u4e2d\u7684":24,"\u51fd\u6570\u53ea\u5173\u6ce8\u4e8ernn\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185\u7684\u8ba1\u7b97":2,"\u51fd\u6570\u5fc5\u987b\u8fd4\u56de\u4e00\u4e2a\u6216\u591a\u4e2alayer\u7684\u8f93\u51fa":2,"\u51fd\u6570\u662f\u4f7f\u7528":24,"\u51fd\u6570\u6765\u4fdd\u8bc1\u517c\u5bb9\u6027":24,"\u51fd\u6570\u67e5\u8be2\u6587\u6863":27,"\u5206\u522b\u4e3a":1,"\u5206\u522b\u4ece\u8bcd\u8bed\u548c\u53e5\u5b50\u7ea7\u522b\u7f16\u7801\u8f93\u5165\u6570\u636e":2,"\u5206\u522b\u5b9a\u4e49\u5b50\u53e5\u7ea7\u522b\u548c\u8bcd\u8bed\u7ea7\u522b\u4e0a\u9700\u8981\u5b8c\u6210\u7684\u8fd0\u7b97":2,"\u5206\u522b\u662f":0,"\u5206\u5e03\u5f0f\u8bad\u7ec3":13,"\u5206\u6790\u5f97\u51fa":1,"\u5206\u7c7b\u6210\u6b63\u9762\u60c5\u7eea\u548c":24,"\u5217\u8868\u5982\u4e0b":24,"\u5219\u53ef\u4ee5\u4f7f\u7528":9,"\u5219\u53ef\u4ee5\u9009\u62e9\u4e0a\u8868\u4e2d\u7684avx\u7248\u672cpaddlepaddl":9,"\u5219\u5b57\u4e0e\u5b57\u4e4b\u95f4\u7528\u7a7a\u683c\u5206\u9694":13,"\u5219\u9700\u8981\u4f7f\u7528":10,"\u5219\u9700\u8981\u4f7f\u7528\u53c2\u6570\u89c4\u6a21\u5927\u7ea65\u500d\u7684\u5185\u5b58":14,"\u5219\u9700\u8981\u5148\u5c06":9,"\u5219\u9700\u8981\u8fdb\u884c\u4e00\u5b9a\u7684\u4e8c\u6b21\u5f00\u53d1":9,"\u521b\u5efa\u4e00\u4e2a":27,"\u521b\u5efagener":24,"\u521d\u59cb\u72b6\u6001":2,"\u5220\u9664contain":9,"\u5229\u7528\u5355\u8bcdid\u67e5\u627e\u5bf9\u5e94\u7684\u8be5\u5355\u8bcd\u7684\u8fde\u7eed\u8868\u793a\u5411\u91cf":13,"\u5229\u7528\u66f4\u591a\u7684\u8ba1\u7b97\u8d44\u6e90\u53ef\u4ee5\u5206\u4e3a\u4e00\u4e0b\u51e0\u4e2a\u65b9\u5f0f\u6765\u8fdb\u884c":14,"\u5229\u7528\u8fd9\u79cd\u7279\u6027":2,"\u5229\u7528\u903b\u8f91\u56de\u5f52\u6a21\u578b\u5bf9\u8be5\u5411\u91cf\u8fdb\u884c\u5206\u7c7b":13,"\u5229\u843d":1,"\u522b":13,"\u5237\u7259":1,"\u524d\u53f0":1,"\u5269\u4e0b\u7684pass\u4f1a\u76f4\u63a5\u4ece\u5185\u5b58\u91cc":24,"\u52a0\u4e86l2\u6b63\u5219\u548c\u68af\u5ea6\u622a\u65ad":13,"\u52a0\u8f7d\u6570\u636e":13,"\u52a0\u901fpaddlepaddle\u8bad\u7ec3\u6709\u5982\u4e0b\u51e0\u4e2a\u65b9\u9762":14,"\u5305":9,"\u5305\u548c":9,"\u5305\u62ec":13,"\u5305\u62ec\u5b57\u7b26\u4e32\u5206\u914d":14,"\u5305\u62ec\u6fc0\u6d3b":14,"\u5305\u62ec\u7b80\u5355\u7684rnn\u6a21\u578b":13,"\u5305\u62ecdocker\u955c\u50cf":8,"\u5305\u62ecpaddle\u7684\u4e8c\u8fdb\u5236":9,"\u5305\u62ecpaddle\u8fd0\u884cdemo\u6240\u9700\u8981\u7684\u4f9d\u8d56":9,"\u5341\u4e00":1,"\u534e\u6da6\u4e07\u5bb6":1,"\u5355\u53cc\u5c42\u5e8f\u5217\u7684\u53e5\u5b50\u662f\u4e00\u6837\u7684":1,"\u5355\u53cc\u5c42\u5e8f\u5217\u7684label\u90fd\u5206\u522b\u662f0\u548c1":1,"\u5355\u5c42":2,"\u5355\u5c42\u5e8f\u5217":[0,1],"\u5355\u5c42\u5e8f\u5217\u7684\u6570\u636e":1,"\u5355\u5c42\u5e8f\u5217\u7684\u6bcf\u4e2a\u5143\u7d20":0,"\u5355\u5c42\u5e8f\u5217\u7684dataprovider\u5982\u4e0b":1,"\u5355\u5c42\u5e8f\u5217\u76f4\u63a5\u53d6\u4e86\u6700\u540e\u4e00\u4e2a\u5143\u7d20":1,"\u5355\u5c42\u5e8f\u5217\u7b2ci\u4e2a\u5143\u7d20":0,"\u5355\u5c42\u5e8f\u5217\u8fc7\u4e86\u4e00\u4e2amix":1,"\u5355\u5c42\u6216\u53cc\u5c42":0,"\u5355\u5c42rnn":[1,2],"\u5355\u5c42rnn\u793a\u4f8b":15,"\u5355\u673acpu\u8bad\u7ec3":14,"\u5355\u673agpu\u8bad\u7ec3":14,"\u5355\u6d4b\u4e2d":1,"\u5355\u8fdb\u5355\u51fa":2,"\u536b\u751f":1,"\u5373":[9,13,14],"\u5373\u4e00\u4e2a\u53e5\u5b50\u4e00\u4e2alabel":1,"\u5373\u4e00\u4e2a\u5b50\u53e5\u4e00\u4e2alabel":1,"\u5373\u4e0d\u5728\u4e4e\u5185\u5b58\u6682\u5b58\u591a\u5c11\u6761\u6570\u636e":24,"\u5373\u4e0d\u662f\u4e00\u6761\u5e8f\u5217":24,"\u5373\u4ece\u5355\u8bcd\u5b57\u7b26\u4e32\u5230\u5355\u8bcdid\u7684\u5b57\u5178":24,"\u5373\u4f1a\u751f\u6210100\u4e2agener":24,"\u5373\u4f7f\u5728check\u4e2d\u6570\u636e\u4e0d\u5408\u6cd5":24,"\u5373\u4f7f\u5728process\u91cc\u9762\u53ea\u4f1a\u6709\u4e00":24,"\u5373\u5185\u5c42memory\u7684boot":1,"\u5373\u521d\u59cb\u72b6\u6001\u4e3a0":2,"\u5373\u5305\u542b\u65f6\u95f4\u6b65\u4fe1\u606f":24,"\u5373\u53cc\u5c42rnn\u7684\u6bcf\u4e2a\u72b6\u6001":2,"\u5373\u53ef":24,"\u5373\u53ef\u4ee5\u4f7f\u7528ssh\u8bbf\u95ee\u5bbf\u4e3b\u673a\u76848022\u7aef\u53e3":9,"\u5373\u53ef\u4ee5\u6781\u5927\u7684\u52a0\u901f\u6570\u636e\u8f7d\u5165\u6d41\u7a0b":14,"\u5373\u53ef\u4ee5\u8bbe\u7f6e\u53c2\u4e0e\u8bad\u7ec3\u7684\u7ebf\u7a0b\u6570\u91cf":14,"\u5373\u53ef\u542f\u52a8\u548c\u8fdb\u5165paddlepaddle\u7684contain":9,"\u5373\u53ef\u5728\u672c\u5730\u7f16\u8bd1\u51fapaddlepaddle\u7684\u955c\u50cf":3,"\u5373\u53ef\u6253\u5370\u51fapaddlepaddle\u7684\u7248\u672c\u548c\u6784\u5efa":9,"\u5373\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d":14,"\u5373\u5927\u90e8\u5206\u503c\u4e3a0":24,"\u5373\u5982\u679ctrain":24,"\u5373\u5bf9\u7b2c3\u6b65\u8fdb\u884c\u66ff\u6362":13,"\u5373\u628a\u5355\u5c42rnn\u751f\u6210\u540e\u7684subseq\u7ed9\u62fc\u63a5\u6210\u4e00\u4e2a\u65b0\u7684\u53cc\u5c42seq":2,"\u5373\u6574\u4e2a\u53cc\u5c42group\u662f\u5c06\u524d\u4e00\u4e2a\u5b50\u53e5\u7684\u6700\u540e\u4e00\u4e2a\u5411\u91cf":1,"\u5373\u6574\u4e2a\u8f93\u5165\u5e8f\u5217":0,"\u5373\u662f\u4e00\u6761\u65f6\u95f4\u5e8f\u5217":24,"\u5373\u8bbe\u7f6e":14,"\u5373\u8d77\u5230\u7684\u4f5c\u7528\u4ec5\u4ec5\u662f\u628a\u53cc\u5c42seq\u62c6\u6210\u5355\u5c42":1,"\u5373input":2,"\u5373train":24,"\u5377\u79ef\u7f51\u7edc\u662f\u4e00\u79cd\u7279\u6b8a\u7684\u4ece\u8bcd\u5411\u91cf\u8868\u793a\u5230\u53e5\u5b50\u8868\u793a\u7684\u65b9\u6cd5":13,"\u53bb\u8fc7":1,"\u53c2\u6570":3,"\u53c2\u6570\u5171\u4eab\u7684\u914d\u7f6e\u793a\u4f8b\u4e3a":14,"\u53c2\u6570\u6570\u91cf":13,"\u53c2\u6570\u6765\u63a7\u5236\u7f13\u5b58\u65b9\u6cd5":14,"\u53c2\u8003":23,"\u53c2\u89c1":[6,7],"\u53c2\u89c1pydataprovider2":1,"\u53c3\u5dee\u7b49\u7b49":14,"\u53c8":1,"\u53c8\u662f\u4e00\u4e2a\u5355\u5c42\u7684\u5e8f\u5217":0,"\u53c8\u8981\u4fdd\u8bc1\u6570\u636e\u662f\u968f\u673a\u7684":14,"\u53cc\u5c42":2,"\u53cc\u5c42\u5e8f\u5217":[0,1],"\u53cc\u5c42\u5e8f\u5217\u5728\u540c\u6837\u7684mix":1,"\u53cc\u5c42\u5e8f\u5217\u662f\u4e00\u4e2a\u5d4c\u5957\u7684\u5e8f\u5217":0,"\u53cc\u5c42\u5e8f\u5217\u662fpaddlepaddle\u652f\u6301\u7684\u4e00\u79cd\u975e\u5e38\u7075\u6d3b\u7684\u6570\u636e\u7ec4\u7ec7\u65b9\u5f0f":2,"\u53cc\u5c42\u5e8f\u5217\u6bcf\u4e2asubseq\u4e2d\u6bcf\u4e2a\u5143\u7d20":0,"\u53cc\u5c42\u5e8f\u5217\u7684\u6570\u636e":1,"\u53cc\u5c42\u5e8f\u5217\u7684dataprovider\u5982\u4e0b":1,"\u53cc\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":0,"\u53cc\u5c42\u5e8f\u5217\u9996\u5148":1,"\u53cc\u5c42\u6216\u8005\u5355\u5c42":0,"\u53cc\u5c42rnn":[1,2],"\u53cc\u5c42rnn\u793a\u4f8b":15,"\u53cc\u8fdb\u5355\u51fa":2,"\u53d1\u884c\u7248":10,"\u53d6\u4e86\u6bcf\u4e2asubseq\u7684\u5e73\u5747\u503c":1,"\u53d6\u4e86\u6bcf\u4e2asubseq\u7684\u6700\u540e\u4e00\u4e2a\u5143\u7d20":1,"\u53d6\u51b3\u4e8e\u662f\u5426\u5bfb\u627e\u5230gflags":4,"\u53d6\u51b3\u4e8e\u662f\u5426\u5bfb\u627e\u5230glog":4,"\u53d6\u51b3\u4e8e\u662f\u5426\u5bfb\u627e\u5230gtest":4,"\u53d6\u51b3\u4e8e\u662f\u5426\u627e\u5230swig":4,"\u53d8\u4e3a3\u4e2a\u65b0\u7684\u5b50\u6b65\u9aa4":13,"\u53d8\u4f1a\u62a5\u8fd9\u4e2a\u9519\u8bef":10,"\u53d8\u91cf":24,"\u53e3\u5934":1,"\u53e5\u5b50\u8868\u793a\u7684\u8ba1\u7b97\u66f4\u65b0\u4e3a2\u6b65":13,"\u53e6":14,"\u53e6\u4e00\u4e2a\u662f\u6bcf\u6761\u5e8f\u5217":14,"\u53ea\u4f5c\u4e3aread":2,"\u53ea\u5305\u62ecpaddle\u7684\u4e8c\u8fdb\u5236":9,"\u53ea\u662f\u53cc\u5c42\u5e8f\u5217\u5c06\u5176\u53c8\u505a\u4e86\u5b50\u5e8f\u5217\u5212\u5206":1,"\u53ea\u662f\u5c06\u53e5\u5b50\u5229\u7528\u8fde\u7eed\u5411\u91cf\u8868\u793a\u66ff\u6362\u7a00\u758f":13,"\u53ea\u662f\u8bf4\u660e\u6570\u636e\u7684\u987a\u5e8f\u662f\u91cd\u8981\u7684":24,"\u53ea\u6709":1,"\u53ea\u7528\u4e8e\u5728\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u4e2d\u6307\u5b9a\u8f93\u5165\u6570\u636e":2,"\u53ea\u80fd\u591f\u8fd4\u56delist\u6216\u8005tupl":24,"\u53ea\u80fd\u901a\u8fc7":1,"\u53ea\u8bfbmemory\u8f93\u5165":2,"\u53ea\u9488\u5bf9\u5185\u5b58":14,"\u53ea\u9700\u8981\u4f7f\u7528\u4e00\u884c\u4ee3\u7801\u5373\u53ef\u4ee5\u8bbe\u7f6e\u8bad\u7ec3\u5f15\u7528\u8fd9\u4e2adataprovid":24,"\u53ea\u9700\u8981\u5728":24,"\u53ea\u9700\u8981\u77e5\u9053\u8fd9\u53ea\u662f\u4e00\u4e2a\u6807\u8bb0\u5c5e\u6027\u7684\u65b9\u6cd5\u5c31\u53ef\u4ee5\u4e86":24,"\u53ef\u4ee5":1,"\u53ef\u4ee5\u4e3a\u4e00\u4e2a\u6570\u636e\u6587\u4ef6\u8fd4\u56de\u591a\u6761\u8bad\u7ec3\u6837\u672c":24,"\u53ef\u4ee5\u4f20\u516510k":3,"\u53ef\u4ee5\u4f30\u8ba1\u51fa\u5982\u679c\u6a21\u578b\u91c7\u7528\u4e0d\u53d8\u7684\u8f93\u51fa\u6700\u5c0f\u7684cost0\u662f\u591a\u5c11":14,"\u53ef\u4ee5\u4f7f\u7528":[3,14],"\u53ef\u4ee5\u4f7f\u7528\u547d\u4ee4":10,"\u53ef\u4ee5\u4f7f\u7528\u5982\u4e0b\u4ee3\u7801":14,"\u53ef\u4ee5\u4f7f\u7528\u8bad\u7ec3\u597d\u7684\u6a21\u578b\u8bc4\u4f30\u5e26\u6709label\u7684\u9a8c\u8bc1\u96c6":13,"\u53ef\u4ee5\u4f7f\u7528graphviz\u5bf9paddlepaddle\u7684\u7f51\u7edc\u6a21\u578b\u8fdb\u884c\u7ed8\u5236":17,"\u53ef\u4ee5\u4f7f\u7528paddl":17,"\u53ef\u4ee5\u4f7f\u7528python\u7684":27,"\u53ef\u4ee5\u51cf\u5c11\u7f13\u5b58\u6c60\u7684\u5927\u5c0f":14,"\u53ef\u4ee5\u53c2\u8003":13,"\u53ef\u4ee5\u5728\u4e00\u4e2a\u51fd\u6570\u91cc":24,"\u53ef\u4ee5\u5728cmake\u7684\u547d\u4ee4\u884c\u8bbe\u7f6e":4,"\u53ef\u4ee5\u5c06\u4e00\u6761\u6570\u636e\u8bbe\u7f6e\u6210\u591a\u4e2abatch":24,"\u53ef\u4ee5\u5c06memory\u7406\u89e3\u4e3a\u4e00\u4e2a\u65f6\u5ef6\u64cd\u4f5c":2,"\u53ef\u4ee5\u5c06paddlepaddle\u7684\u6a21\u578b\u548c\u914d\u7f6e\u6253\u5305\u6210\u4e00\u4e2a\u6587\u4ef6":17,"\u53ef\u4ee5\u5c06paddlepaddle\u7684\u8bad\u7ec3\u6a21\u578b\u4ee5proto":17,"\u53ef\u4ee5\u65b9\u4fbf\u5d4c\u5165\u5f0f\u5de5\u4f5c":4,"\u53ef\u4ee5\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":2,"\u53ef\u4ee5\u662f\u4e00\u4e2a\u975e\u5e8f\u5217":2,"\u53ef\u4ee5\u663e\u793a\u5730\u6307\u5b9a\u4e00\u4e2alayer\u7684\u8f93\u51fa\u7528\u4e8e\u521d\u59cb\u5316memori":2,"\u53ef\u4ee5\u6709\u4ee5\u4e0b\u4e24\u79cd":2,"\u53ef\u4ee5\u6839\u636e\u4e0d\u540c\u7684\u6570\u636e\u914d\u7f6e\u4e0d\u540c\u7684\u8f93\u5165\u7c7b\u578b":24,"\u53ef\u4ee5\u770b\u4f5c\u662f\u4e00\u4e2a\u975e\u5e8f\u5217\u8f93\u5165":0,"\u53ef\u4ee5\u8003\u8651\u4f7f\u7528\u4e00\u4e9b\u4f18\u5316\u7b97\u6cd5":14,"\u53ef\u4ee5\u8fd4\u56de\u4e00\u4e2adict":24,"\u53ef\u4ee5\u901a\u8fc7show":13,"\u53ef\u7528\u5728\u6d4b\u8bd5\u6216\u8bad\u7ec3\u65f6\u6307\u5b9a\u521d\u59cb\u5316\u6a21\u578b":13,"\u53ef\u80fd\u7684\u4ee3\u7801\u4e3a":14,"\u53ef\u80fd\u7684\u5185\u5b58\u6cc4\u9732\u95ee\u9898":23,"\u53ef\u80fd\u7684\u8f93\u51fa\u4e3a":10,"\u53ef\u9009":24,"\u5403":1,"\u5403\u996d":1,"\u5404\u65b9\u9762":1,"\u5404\u79cd\u53c2\u6570\u548c\u7ef4\u62a4":3,"\u5408":1,"\u5408\u7406":1,"\u540c\u65f6":[3,4,14,24],"\u540c\u65f6\u4e5f\u53ef\u4ee5\u52a0\u901f\u5f00\u59cb\u8bad\u7ec3\u524d\u6570\u636e\u8f7d\u5165\u7684\u8fc7\u7a0b":14,"\u540c\u65f6\u4e5f\u80fd\u591f\u5f15\u5165\u66f4\u52a0\u590d\u6742\u7684\u8bb0\u5fc6\u673a\u5236":2,"\u540c\u65f6\u4f1a\u8ba1\u7b97\u5206\u7c7b\u51c6\u786e\u7387":13,"\u540c\u65f6\u5728\u5185\u5b58\u91cc\u76f4\u63a5\u968f\u5373\u9009\u53d6\u6570\u636e\u6765\u505ashuffl":14,"\u540c\u65f6\u5c06\u53c2\u6570\u521d\u59cb\u5316\u4e3a":14,"\u540c\u65f6\u6b22\u8fce\u8d21\u732e\u66f4\u591a\u7684\u5b89\u88c5\u5305":8,"\u540c\u65f6\u8bbe\u7f6e\u5185\u5b58\u7f13\u5b58\u529f\u80fd":14,"\u540c\u6837\u53ef\u4ee5\u6269\u5c55\u5230\u53cc\u5c42\u5e8f\u5217\u7684\u5904\u7406\u4e0a":2,"\u540d\u79f0":13,"\u540e":14,"\u540e\u9762\u8ddf\u7740\u4e00\u7cfb\u5217\u7f16\u8bd1\u53c2\u6570":22,"\u5411\u91cf\u8868\u793a":13,"\u5426":4,"\u5426\u5219":23,"\u5426\u5219\u5728\u7b2c0\u4e2a\u65f6\u95f4\u6b65\u65f6":1,"\u5426\u5219\u9700\u8981\u9009\u62e9\u975eavx\u7684paddlepaddl":9,"\u5440":1,"\u5468\u56f4":1,"\u547d\u4ee4":3,"\u547d\u4ee4\u4e3a":9,"\u547d\u4ee4\u5373\u53ef\u5b8c\u6210\u5b89\u88c5":10,"\u547d\u4ee4\u6307\u5b9a\u7684\u53c2\u6570\u4f1a\u4f20\u5165\u7f51\u7edc\u914d\u7f6e\u4e2d":13,"\u547d\u4ee4\u8fd0\u884c\u955c\u50cf":9,"\u547d\u4ee4\u9884\u5148\u4e0b\u8f7d\u955c\u50cf":9,"\u548c\u4e00\u4e2a\u5df2\u7ecf\u5206\u8bcd\u540e\u7684\u53e5\u5b50":1,"\u548c\u4e09\u79cd\u5e8f\u5217\u6a21\u5f0f":24,"\u548c\u4e4b\u524d\u51cf\u5c0f\u901a\u8fc7\u51cf\u5c0f\u7f13\u5b58\u6c60\u6765\u51cf\u5c0f\u5185\u5b58\u5360\u7528\u7684\u539f\u7406\u4e00\u81f4":14,"\u548c\u53cc\u5c42\u5e8f\u5217\u542b\u6709subseq":0,"\u548c\u53cc\u5c42rnn":1,"\u548c\u5dee\u8bc4":13,"\u548c\u5e8f\u5217\u4e2d\u542b\u6709\u5143\u7d20\u7684\u6570\u76ee\u540c":0,"\u548c\u6587\u672c\u4fe1\u606f\u7528tab\u95f4\u9694":13,"\u548c\u6d4b\u8bd5\u6587\u4ef6\u5217\u8868":23,"\u548c\u7528\u6237\u4f20\u5165\u7684\u53c2\u6570":24,"\u548c\u90e8\u5206layer":2,"\u548c\u9884\u5904\u7406\u811a\u672c":13,"\u548cavgpool":0,"\u548ccudnn":10,"\u548cinitalizer\u91cc\u5b9a\u4e49\u987a\u5e8f\u4e00\u81f4":13,"\u54c1\u8d28":1,"\u5546\u52a1":1,"\u554a":1,"\u5668":13,"\u56db\u4e2a\u7248\u672c":10,"\u56db\u79cd\u6570\u636e\u7c7b\u578b\u662f":24,"\u56e0\u6b64":2,"\u56e0\u6b642\u4e2abatch\u5c31\u53ef\u4ee5\u5b8c\u62101\u4e2apass":1,"\u56e0\u6b64\u4e0a\u8ff0\u4e09\u4e2alayer\u7684\u524d\u5411\u4f1a\u62a5\u51fa":1,"\u56e0\u6b64\u4e24\u4e2a\u914d\u7f6e\u5728\u8fd9\u4e24\u5c42\u4e0a\u7684\u8f93\u51fa\u662f\u4e00\u6837\u7684":1,"\u56e0\u6b64\u5355\u53cc\u5c42\u5e8f\u5217\u7684\u8f93\u51fa\u662f\u4e00\u6837\u65f3":1,"\u56e0\u6b64\u53cc\u5c42\u5e8f\u5217\u7684\u914d\u7f6e\u4e2d":1,"\u56e0\u6b64\u53cc\u5c42\u5e8f\u5217\u8fc7\u5b8clstmemory\u7684\u8f93\u51fa\u548c\u5355\u5c42\u7684\u4e00\u6837":1,"\u56e0\u6b64\u5f53\u5916\u5c42\u6709i":1,"\u56fe\u50cf\u5206\u7c7b":12,"\u5728":[0,1,4,10,24],"\u5728\u4e4b\u540e\u7684":14,"\u5728\u4f7f\u7528":14,"\u5728\u53cd\u5411\u4f20\u9012\u7684\u65f6\u5019":14,"\u5728\u58f0\u660edataprovider\u7684\u65f6\u5019\u4f20\u5165\u4e86dictionary\u4f5c\u4e3a\u53c2\u6570":24,"\u5728\u5b8c\u6210\u4e86\u6570\u636e\u548c\u7f51\u7edc\u7ed3\u6784\u642d\u5efa\u4e4b\u540e":13,"\u5728\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u4e2d":2,"\u5728\u672c\u95ee\u9898\u4e2d":13,"\u5728\u6a21\u578b\u914d\u7f6e\u4e2d\u5229\u7528":13,"\u5728\u6b64\u4e3a\u65b9\u4fbf\u5bf9\u6bd4\u4e0d\u540c\u7f51\u7edc\u7ed3\u6784":13,"\u5728\u6bcf\u4e2a\u7ef4\u5ea6\u4e0a\u53d6\u51fa\u5728\u8be5\u53e5\u8bdd\u65b0\u7684\u5411\u91cf\u96c6\u5408\u4e0a\u8be5\u7ef4\u5ea6\u7684\u6700\u5927\u503c\u4f5c\u4e3a\u6700\u540e\u7684\u53e5\u5b50\u8868\u793a\u5411\u91cf":13,"\u5728\u7a0b\u5e8f\u5f00\u59cb\u9636\u6bb5":27,"\u5728\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4efb\u52a1\u4e2d":0,"\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u8fdb\u884c\u6d4b\u8bd5":23,"\u5728\u8bad\u7ec3\u914d\u7f6e\u91cc":24,"\u5728\u8f93\u51fa\u7684\u8fc7\u7a0b\u4e2d":2,"\u5728\u8fd9\u4e2a\u51fd\u6570\u4e2d":24,"\u5728\u8fd9\u79cd\u7ed3\u6784\u4e2d":2,"\u5728\u8fd9\u91cc":2,"\u5728\u914d\u7f6e\u4e2d\u8bfb\u53d6\u4e86\u5b57\u5178":24,"\u5728\u91c7\u7528sgd":14,"\u5728cmake\u914d\u7f6e\u65f6\u53ef\u4ee5\u4f7f\u7528":4,"\u5728paddlepaddle\u4e2d":2,"\u5728pydataprovider\u4e2d":24,"\u5728python\u73af\u5883\u4e0b\u9884\u6d4b\u7ed3\u679c":27,"\u5728step\u51fd\u6570\u4e2d\u5b9a\u4e49":2,"\u5728step\u51fd\u6570\u4e2d\u5b9a\u4e49memori":2,"\u5730\u6bb5":1,"\u5730\u7406\u4f4d\u7f6e":1,"\u5730\u94c1\u7ad9":1,"\u5747\u5300\u5206\u5e03":14,"\u57fa\u4e8e\u53cc\u5c42\u5e8f\u5217\u8f93\u5165":2,"\u57fa\u672c\u4e0a\u4e0d\u80fd\u6574\u4f53\u4fee\u6b63":24,"\u57fa\u672c\u7684\u5904\u7406\u903b\u8f91\u4e5f\u548cmnist\u903b\u8f91\u4e00\u81f4":24,"\u57fa\u672c\u7684pydataprovider\u4f7f\u7528\u4ecb\u7ecd\u5b8c\u6bd5\u4e86":24,"\u5904\u7406\u7684\u8f93\u5165\u5e8f\u5217\u4e3b\u8981\u5206\u4e3a\u4ee5\u4e0b\u4e09\u79cd\u7c7b\u578b":2,"\u5916\u5c42memory\u5fc5\u987b\u6709boot":1,"\u5916\u5c42memory\u662f\u4e00\u4e2a\u5143\u7d20":1,"\u5916\u5c42memory\u662f\u5355\u5c42\u5e8f\u5217":1,"\u5916\u5c42outer":1,"\u591a\u4e2ainput\u4ee5list\u65b9\u5f0f\u8f93\u5165":13,"\u591a\u53e5\u8bdd\u8fdb\u4e00\u6b65\u6784\u6210\u4e86\u6bb5\u843d":2,"\u591a\u673a\u8bad\u7ec3":14,"\u591a\u6b21\u8fd4\u56de\u53d8\u91cf":24,"\u591a\u7ebf\u7a0b\u4e0b\u8f7d\u8fc7\u7a0b\u4e2d":3,"\u591a\u7ebf\u7a0b\u6570\u636e\u8bfb\u53d6":24,"\u591a\u8f6e\u5bf9\u8bdd\u7b49\u66f4\u4e3a\u590d\u6742\u7684\u8bed\u8a00\u6570\u636e":2,"\u5927":24,"\u5929":1,"\u5929\u4e00\u5e7f\u573a":1,"\u5929\u4e00\u9601":1,"\u597d":1,"\u597d\u5403":1,"\u597d\u8bc4":13,"\u5982\u4e0b":1,"\u5982\u679c":[10,24],"\u5982\u679c\u4e0d\u4e86\u89e3":24,"\u5982\u679c\u4e0d\u4f7f\u7528\u5219\u4f1a\u4f7f\u7528\u4e00\u4e2a\u7b80\u5316\u7248\u7684\u547d\u4ee4\u884c\u53c2\u6570\u89e3\u6790":4,"\u5982\u679c\u4e0d\u4f7f\u7528\u5219\u4f1a\u4f7f\u7528\u4e00\u4e2a\u7b80\u5316\u7248\u7684\u65e5\u5fd7\u5b9e\u73b0":4,"\u5982\u679c\u4e0d\u5207\u8bcd":13,"\u5982\u679c\u4e0d\u6536\u655b":14,"\u5982\u679c\u4e0d\u8bbe\u7f6e\u7684\u8bdd":24,"\u5982\u679c\u4f7f\u7528gpu\u7248\u672c\u7684paddlepaddl":10,"\u5982\u679c\u5185\u5c42memory\u60f3":1,"\u5982\u679c\u53c2\u6570\u4fdd\u5b58\u4e0b\u6765\u7684":14,"\u5982\u679c\u5728":10,"\u5982\u679c\u5728\u7b2c\u4e00\u6b21cmake\u4e4b\u540e\u60f3\u8981\u91cd\u65b0\u8bbe":4,"\u5982\u679c\u5728\u8bad\u7ec3\u65f6":24,"\u5982\u679c\u5c06\u8fd9\u4e2a\u5185\u5b58\u6c60\u51cf\u5c0f":14,"\u5982\u679c\u5c0f\u4e8e\u8fd9\u4e2a\u4e0b\u8f7d\u901f\u5ea6":3,"\u5982\u679c\u60a8\u4f7f\u7528":9,"\u5982\u679c\u60f3\u8981\u5728\u5916\u90e8\u673a\u5668\u8bbf\u95ee\u8fd9\u4e2acontain":9,"\u5982\u679c\u662ffalse\u7684\u8bdd":24,"\u5982\u679c\u662ftrue\u7684\u8bdd":24,"\u5982\u679c\u6709\u591a\u4e2a\u8f93\u5165":2,"\u5982\u679c\u6709\u591a\u4e2a\u8f93\u5165\u5e8f\u5217":2,"\u5982\u679c\u6709\u66f4\u590d\u6742\u7684\u4f7f\u7528":23,"\u5982\u679c\u6ca1\u6709\u5b9a\u4e49memori":2,"\u5982\u679c\u7528\u6237\u4e0d\u6307\u5b9a\u8fd4\u56de\u6570\u636e\u7684\u5bf9\u5e94\u5173\u7cfb":24,"\u5982\u679c\u7528\u6237\u60f3\u8981\u81ea\u5b9a\u4e49\u521d\u59cb\u5316\u65b9\u5f0f":14,"\u5982\u679c\u8bad\u7ec3\u4e00\u4e2apass":14,"\u5982\u679c\u8bad\u7ec3\u8fc7\u7a0b\u7684\u7684cost\u660e\u663e\u9ad8\u4e8e\u8fd9\u4e2a\u5e38\u6570\u8f93\u51fa\u7684cost":14,"\u5982\u679c\u8bbe\u7f6e\u6210true\u7684\u8bdd":24,"\u5982\u679c\u8f93\u51fa":9,"\u5982\u679c\u8fd0\u884cgpu\u7248\u672c\u7684paddlepaddl":9,"\u5982\u679clearning_rate\u592a\u5927":14,"\u5982\u679clearning_rate\u592a\u5c0f":14,"\u5982\u679ctest":23,"\u5b50":1,"\u5b50\u53e5":2,"\u5b50\u53e5\u7684\u5355\u8bcd\u6570\u548c\u6307\u5b9a\u7684\u4e00\u4e2a\u8f93\u5165\u5e8f\u5217\u4e00\u81f4":2,"\u5b81\u6ce2":1,"\u5b83\u5305\u542b\u7684\u53c2\u6570\u6709":24,"\u5b83\u7684":1,"\u5b83\u7684\u6bcf\u4e00\u4e2a\u5143\u7d20":0,"\u5b89\u6392":1,"\u5b89\u88c5\u5305":14,"\u5b89\u88c5\u5305\u5728ubuntu":10,"\u5b89\u88c5\u5305\u7684\u4e0b\u8f7d\u5730\u5740\u662f":10,"\u5b89\u88c5\u597d\u7684paddlepaddle\u811a\u672c\u5305\u62ec\u591a\u6761\u547d\u4ee4":17,"\u5b89\u88c5\u5b8c\u6210\u540e":10,"\u5b89\u88c5\u5b8c\u6210\u7684paddlepaddle\u4e3b\u4f53\u5305\u62ec\u4e09\u4e2a\u90e8\u5206":9,"\u5b89\u88c5\u5b8c\u6210paddlepaddle\u540e":10,"\u5b89\u88c5\u6559\u7a0b":13,"\u5b89\u88c5\u65b9\u6cd5\u8bf7\u53c2\u8003":9,"\u5b89\u88c5\u7f16\u8bd1\u4f9d\u8d56":6,"\u5b89\u88c5\u7f16\u8bd1paddlepaddle\u9700\u8981\u7684\u4f9d\u8d56":5,"\u5b89\u88c5docker\u9700\u8981\u60a8\u7684\u673a\u5668":9,"\u5b89\u88c5paddlepaddl":13,"\u5b89\u88c5paddlepaddle\u7684docker\u955c\u50cf":8,"\u5b89\u9759":1,"\u5b8c\u6210\u4efb\u610f\u7684\u8fd0\u7b97\u903b\u8f91":2,"\u5b8c\u6210\u591a\u673a\u8bad\u7ec3":17,"\u5b8c\u6210\u76f8\u5e94\u7684\u8ba1\u7b97":0,"\u5b8c\u6574\u4ee3\u7801\u89c1":27,"\u5b9a\u4e49\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185rnn\u5355\u5143\u5b8c\u6210\u7684\u8ba1\u7b97":2,"\u5b9a\u4e49\u4e86\u4e00\u4e2a\u53ea\u8bfb\u7684memori":2,"\u5b9a\u4e49\u5728\u5916\u5c42":2,"\u5b9a\u4e49\u6587\u672c\u4fe1\u606f":13,"\u5b9e\u73b0\u4e86\u6253\u5f00\u6587\u672c\u6587\u4ef6":24,"\u5b9e\u73b0\u8bcd\u8bed\u548c\u53e5\u5b50\u4e24\u4e2a\u7ea7\u522b\u7684\u53cc\u5c42rnn\u7ed3\u6784":2,"\u5b9e\u9645\u4e2d\u5e76\u4e0d\u9700\u8981":1,"\u5ba2\u6237":1,"\u5bb6":1,"\u5bc6\u7801\u4e5f\u662froot":9,"\u5bf9":1,"\u5bf9\u4e8e\u7528\u6237\u6765\u8bf4":24,"\u5bf9\u4e8e\u7ed9\u5b9a\u7684\u4e00\u6761\u6587\u672c":13,"\u5bf9\u4e8ecuda\u7684toolkit\u6709\u65ad\u70b9\u7eed\u4f20\u548c\u4f20\u8f93\u901f\u5ea6\u8fc7\u5c0f\u91cd\u542f\u4e0b\u8f7d\u7684":3,"\u5bf9\u4e8emnist\u800c\u8a00":24,"\u5bf9\u5e94\u4e00\u4e2a\u5b50\u53e5":2,"\u5bf9\u5e94\u4e00\u4e2a\u8bcd":2,"\u5bf9\u8be5\u8868\u793a\u8fdb\u884c\u975e\u7ebf\u6027\u53d8\u6362":13,"\u5bf9\u8c61":[14,24],"\u5bf9\u8c61convert":27,"\u5bf9\u8f93\u51fa\u7684\u5408\u5e76":2,"\u5bf9\u9762":1,"\u5bfc\u81f4\u8bad\u7ec3\u65f6\u95f4\u8fc7\u957f":14,"\u5c06":14,"\u5c06\u4f1a\u6d4b\u8bd5\u914d\u7f6e\u6587\u4ef6\u4e2dtest":13,"\u5c06\u5176\u6269\u5c55\u6210\u4e00\u4e2a\u65b0\u7684\u53cc\u5c42\u5e8f\u5217":1,"\u5c06\u5176\u62fc\u63a5\u6210\u4e00\u4e2a\u65b0\u7684\u5355\u5c42\u5e8f\u5217":1,"\u5c06\u5176\u8bbe\u7f6e\u6210":14,"\u5c06\u542b\u6709\u5b50\u53e5":2,"\u5c06\u542b\u6709\u8bcd\u8bed\u7684\u53e5\u5b50\u5b9a\u4e49\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":2,"\u5c06\u5b57\u5178\u5b58\u5165\u4e86set":24,"\u5c06\u5bbf\u4e3b\u673a\u76848022\u7aef\u53e3\u6620\u5c04\u5230container\u768422\u7aef\u53e3\u4e0a":9,"\u5c06\u6570\u636e\u7ec4\u5408\u6210batch\u8bad\u7ec3":24,"\u5c06\u6587\u4ef6\u7684\u7edd\u5bf9\u8def\u5f84\u6216\u76f8\u5bf9\u8def\u5f84":23,"\u5c06\u8bc4\u8bba\u5206\u4e3a\u597d\u8bc4":13,"\u5c06\u8be5\u53e5\u8bdd\u5305\u542b\u7684\u6240\u6709\u5355\u8bcd\u5411\u91cf\u6c42\u5e73\u5747\u5f97\u5230\u53e5\u5b50\u7684\u8868\u793a":13,"\u5c06ssh\u88c5\u5165\u7cfb\u7edf\u5185\u5e76\u5f00\u542f\u8fdc\u7a0b\u8bbf\u95ee":9,"\u5c1a\u53ef":1,"\u5c31":[1,24],"\u5c31\u50cf\u8fd9\u4e2a\u6837\u4f8b\u4e00\u6837":24,"\u5c31\u5f88\u5bb9\u6613\u5bfc\u81f4\u5185\u5b58\u8d85\u9650":14,"\u5c31\u662f":1,"\u5c31\u662f\u5c06\u8fd9\u4e9b\u52a8\u6001\u5e93\u52a0\u5230\u73af\u5883\u53d8\u91cf\u91cc\u9762":10,"\u5c42\u6b21\u5316\u7684rnn":2,"\u5c45\u7136":1,"\u5c5e\u6027":24,"\u5dee\u8bc4":13,"\u5e2e\u52a9\u6211\u4eec\u5b8c\u6210\u5bf9\u8f93\u5165\u5e8f\u5217\u7684\u62c6\u5206":2,"\u5e2e\u52a9\u6211\u4eec\u66f4\u597d\u5730\u63cf\u8ff0\u6bb5\u843d":2,"\u5e2e\u52a9\u6211\u4eec\u6784\u9020\u4e00\u4e9b\u590d\u6742\u7684\u8f93\u5165\u4fe1\u606f":0,"\u5e38\u89c1\u7684\u8f93\u51fa\u683c\u5f0f\u4e3a":22,"\u5e72\u51c0":1,"\u5e76\u4e14":24,"\u5e76\u4e14\u4f7f\u7528\u5173\u952e\u8bcd":24,"\u5e76\u4e14\u5220\u9664container\u4e2d\u7684\u6570\u636e":9,"\u5e76\u4e14\u53ef\u80fd\u4f1a\u52a0\u901f\u8bad\u7ec3\u8fc7\u7a0b":14,"\u5e76\u4e14\u5728\u5185\u5b58\u8db3\u591f":24,"\u5e76\u4e14\u6807\u8bb0process\u51fd\u6570\u662f\u4e00\u4e2adataprovid":24,"\u5e76\u4e14softmax\u5c42\u7684\u4e24\u4e2a\u8f93\u5165\u4e5f\u4f7f\u7528\u4e86\u540c\u6837\u7684\u53c2\u6570":14,"\u5e76\u4f7f\u7528\u4e86dropout":13,"\u5e76\u572823\u884c\u8fd4\u56de\u7ed9paddlepaddle\u8fdb\u7a0b":24,"\u5e76\u5bf9\u5176\u8be6\u7ec6\u5206\u6790":1,"\u5e76\u5c06\u6bcf\u884c\u8f6c\u6362\u6210\u548c":24,"\u5e76\u63d0\u4f9b":9,"\u5e76\u63d0\u4f9b\u4e86\u7b80\u5355\u7684cache\u529f\u80fd":24,"\u5e76\u8bbe\u7f6e\u597d\u5bf9\u5e94\u7684\u73af\u5883\u53d8\u91cf":10,"\u5e76\u9010\u6e10\u5c55\u793a\u66f4\u52a0\u6df1\u5165\u7684\u529f\u80fd":13,"\u5e8a\u4e0a\u7528\u54c1":1,"\u5e8a\u57ab":1,"\u5e8f\u5217\u4e2d\u542b\u6709\u5143\u7d20\u7684\u6570\u76ee\u540clayer2\u4e00\u81f4":0,"\u5e8f\u5217\u6570\u636e\u548c\u4e0a\u9762\u7684\u5b8c\u5168\u4e00\u6837":1,"\u5e8f\u5217\u6570\u636e\u662f\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4efb\u52a1\u9762\u5bf9\u7684\u4e00\u79cd\u4e3b\u8981\u8f93\u5165\u6570\u636e\u7c7b\u578b":2,"\u5e8f\u5217\u662f\u4e00\u79cd\u5e38\u89c1\u7684\u6570\u636e\u7c7b\u578b":0,"\u5e8f\u5217\u6a21\u578b\u6570\u636e\u63d0\u4f9b":23,"\u5e8f\u5217\u6a21\u578b\u662f\u6307\u6570\u636e\u7684\u67d0\u4e00\u7ef4\u5ea6\u662f\u4e00\u4e2a\u5e8f\u5217\u5f62\u5f0f":24,"\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u5927\u591a\u9075\u5faaencod":2,"\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u7684\u8f93\u5165":2,"\u5e8f\u5217\u7684\u6bcf\u4e2a\u5143\u7d20\u662f\u539f\u6765\u53cc\u5c42\u5e8f\u5217\u6bcf\u4e2asubseq\u5143\u7d20\u7684\u5e73\u5747\u503c":0,"\u5e93\u7684\u8bdd":10,"\u5e94\u8be5":1,"\u5e94\u8be5\u964d\u4f4e\u5b66\u4e60\u7387":14,"\u5f00\u542fgpu\u8bad\u7ec3":14,"\u5f0f":24,"\u5f15\u7528\u7684dataprovider\u662f":24,"\u5f15\u7528memory\u5f97\u5230\u8fd9layer\u4e0a\u4e00\u65f6\u523b\u8f93\u51fa":2,"\u5f3a\u70c8\u63a8\u8350":1,"\u5f53\u51fd\u6570\u8fd4\u56de\u7684\u65f6\u5019":24,"\u5f53\u524d\u7684\u8f93\u5165y\u548c\u4e0a\u4e00\u4e2a\u65f6\u95f4\u6b65\u7684\u8f93\u51farnn":1,"\u5f53\u524dlog_period\u4e2abatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747\u5206\u7c7b\u9519\u8bef\u7387":13,"\u5f53\u524dlog_period\u4e2abatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747cost":13,"\u5f53\u7136":23,"\u5f53\u8c03":24,"\u5f62\u6210recurr":2,"\u5f62\u6210recurrent\u8fde\u63a5":2,"\u5f88":[1,13],"\u5f88\u591a":1,"\u5f88\u5b89\u9759":1,"\u5f88\u5e72\u51c0":1,"\u5f88\u65b9\u4fbf":1,"\u5f97":1,"\u5f97\u5230\u7ed3\u679c":10,"\u5faa\u73af\u5c55\u5f00\u7684\u6bcf\u4e2a\u65f6\u95f4\u6b65\u603b\u662f\u80fd\u591f\u5f15\u7528\u6240\u6709\u8f93\u5165":2,"\u5fc5\u987b\u5c06\u524d\u4e00\u4e2a\u5b50\u53e5\u7684\u6700\u540e\u4e00\u4e2a\u5143\u7d20":1,"\u5fc5\u987b\u6307\u5411\u4e00\u4e2apaddlepaddle\u5b9a\u4e49\u7684lay":2,"\u5fc5\u987b\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":2,"\u5fc5\u987b\u662f\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":0,"\u5feb":1,"\u5feb\u901f\u5165\u95e8":15,"\u5ff5\u662f":24,"\u6027\u4ef7\u6bd4":1,"\u603b\u4f53\u6765\u8bf4":1,"\u60a8\u4e5f\u53ef\u4ee5\u91c7\u7528\u522b\u7684\u7ec4\u7ec7\u5f62\u5f0f":1,"\u60a8\u53ef\u4ee5\u4f7f\u7528":9,"\u60a8\u5c31\u53ef\u4ee5\u8fdc\u7a0b\u7684\u4f7f\u7528paddlepaddle\u5566":9,"\u60a8\u9700\u8981\u5728\u673a\u5668\u4e2d\u5b89\u88c5\u597ddocker":9,"\u60a8\u9700\u8981\u8fdb\u5165\u955c\u50cf\u8fd0\u884cpaddlepaddl":9,"\u60c5\u611f\u5206\u6790":12,"\u60f3\u8981\u8fd0\u884cpaddlepaddl":9,"\u611f\u89c9":1,"\u6210\u4e3a\u7ef4\u5ea6\u4e3ahidden":13,"\u6211\u4eec\u4ece\u63d0\u524d\u7ed9\u5b9a\u7684\u7c7b\u522b\u96c6\u5408\u4e2d\u9009\u62e9\u5176\u6240\u5c5e\u7c7b":13,"\u6211\u4eec\u4ee5\u6587\u672c\u5206\u7c7b\u95ee\u9898\u4f5c\u4e3a\u80cc\u666f":13,"\u6211\u4eec\u4f7f\u7528":13,"\u6211\u4eec\u53ef\u4ee5\u6309\u7167\u5982\u4e0b\u5c42\u6b21\u5b9a\u4e49\u975e\u5e8f\u5217":0,"\u6211\u4eec\u53ef\u4ee5\u8bbe\u8ba1\u642d\u5efa\u4e00\u4e2a\u7075\u6d3b\u7684":2,"\u6211\u4eec\u5728":1,"\u6211\u4eec\u5728\u6b64\u603b":13,"\u6211\u4eec\u5c06\u4ee5\u57fa\u672c\u7684\u903b\u8f91\u56de\u5f52\u7f51\u7edc\u4f5c\u4e3a\u8d77\u70b9":13,"\u6211\u4eec\u5c06\u5728\u540e\u9762\u4ecb\u7ecd\u8bad\u7ec3\u548c\u9884\u6d4b\u7684\u6d41\u7a0b\u7684\u811a\u672c":13,"\u6211\u4eec\u5c06\u8bad\u7ec3\u7684\u547d\u4ee4\u884c\u4fdd\u5b58\u5728\u4e86":13,"\u6211\u4eec\u63a8\u8350\u4f7f\u7528docker\u955c\u50cf\u6765\u90e8\u7f72\u73af\u5883":8,"\u6211\u4eec\u63d0\u4f9b\u4e8612\u4e2a":9,"\u6211\u4eec\u63d0\u4f9b\u4e86\u4e00\u4e2a\u5de5\u5177\u7c7bdataproviderconvert":27,"\u6211\u4eec\u770b\u4e00\u4e0b\u5355\u53cc\u5c42\u5e8f\u5217\u7684\u4e0d\u540c\u6570\u636e\u7ec4\u7ec7\u5f62\u5f0f":1,"\u6211\u4eec\u770b\u4e00\u4e0b\u5355\u53cc\u5c42\u5e8f\u5217\u7684\u4e0d\u540c\u6570\u636e\u7ec4\u7ec7\u5f62\u5f0f\u548cdataprovid":1,"\u6211\u4eec\u770b\u4e00\u4e0b\u5355\u53cc\u5c42\u5e8f\u5217\u7684\u4e0d\u540cdataprovid":1,"\u6211\u4eec\u770b\u4e00\u4e0b\u5355\u5c42\u5e8f\u5217\u7684\u914d\u7f6e":1,"\u6211\u4eec\u770b\u4e00\u4e0b\u8bed\u4e49\u76f8\u540c\u7684\u53cc\u5c42\u5e8f\u5217\u914d\u7f6e":1,"\u6211\u4eec\u79f0\u4e4b\u4e3a\u4e00\u4e2a0\u5c42\u7684\u5e8f\u5217":0,"\u6211\u4eec\u8fdb\u5165\u5230\u8bad\u7ec3\u90e8\u5206":13,"\u6211\u4eec\u9009\u53d6\u5355\u53cc\u5c42\u5e8f\u5217\u914d\u7f6e\u4e2d\u7684\u4e0d\u540c\u90e8\u5206":1,"\u6211\u4eec\u91c7\u7528\u5355\u5c42lstm\u6a21\u578b":13,"\u6211\u4eec\u968f\u65f6\u603b\u7ed3\u4e86\u5404\u4e2a\u7f51\u7edc\u7684\u590d\u6742\u5ea6\u548c\u6548\u679c":13,"\u6216":1,"\u6216\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"\u6216\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":0,"\u6216\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":0,"\u6216\u4e00\u4e2a\u5411\u91cf":2,"\u6216\u5176\u4ed6":13,"\u6216\u5355\u5c42\u5e8f\u5217":0,"\u6216\u5355\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":0,"\u6216\u6700\u5927\u503c":0,"\u6216\u7b2c\u4e00\u4e2a":0,"\u6216\u7b2c\u4e00\u4e2a\u5143\u7d20":0,"\u6216\u8005":[0,9,14],"\u6216\u800510g\u8fd9\u6837\u7684\u5355\u4f4d":3,"\u6216\u8005\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":2,"\u6216\u8005\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":[0,2],"\u6216\u8005\u4f7f\u7528\u4e0b\u9762\u4e00\u6761\u547d\u4ee4\u5b89\u88c5":10,"\u6216\u8005\u5728python":24,"\u6216\u8005\u6570\u636e\u5e93\u8fde\u63a5\u5730\u5740\u7b49\u7b49":23,"\u6216\u8005\u662f\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":0,"\u6216\u8005\u66f4\u65e9":14,"\u6216\u8005\u76f4\u63a5\u6254\u6389\u975e\u5e38\u957f\u7684\u5e8f\u5217":14,"\u6216\u8005\u8bbe\u7f6e\u4e3anone":23,"\u6216\u8005\u9700\u8981\u66f4\u9ad8\u7684\u6548\u7387":23,"\u6216\u8005\u9ad8\u6027\u80fd\u7684":9,"\u623f":1,"\u623f\u95f4":1,"\u6240\u4ee5":[14,24,27],"\u6240\u4ee5\u5728cpu\u7684\u8fd0\u7b97\u6027\u80fd\u4e0a\u5e76\u4e0d\u4f1a\u6709\u4e25\u91cd\u7684\u5f71\u54cd":9,"\u6240\u4ee5\u5982\u679c\u5bf9\u4e8e\u5185\u5b58\u6bd4\u8f83\u5c0f\u7684\u673a\u5668":24,"\u6240\u4ee5\u5982\u679c\u60f3\u8981\u5728\u540e\u53f0\u542f\u7528ssh":9,"\u6240\u4ee5\u5c06":24,"\u6240\u4ee5\u63a8\u8350\u4f7f\u7528\u663e\u5f0f\u6307\u5b9a\u8fd4\u56de\u503c\u548c\u6570\u636e\u5bf9\u5e94\u5173\u7cfb":24,"\u6240\u4ee5\u6700\u4f73\u5b9e\u8df5\u63a8\u8350\u4e0d\u8981\u5c06\u6bcf\u4e00\u4e2a\u6837\u672c\u90fd\u653e\u5165train":24,"\u6240\u4ee5\u7528\u4e8e\u9884\u6d4b\u7684\u914d\u7f6e\u6587\u4ef6\u8981\u505a\u76f8\u5e94\u7684\u4fee\u6539":27,"\u6240\u4ee5\u8f93\u51fa\u7684value\u5305\u542b\u4e24\u4e2a\u5411\u91cf":27,"\u6240\u4ee5gpu\u5728\u8fd0\u7b97\u6027\u80fd\u4e0a\u4e5f\u4e0d\u4f1a\u6709\u4e25\u91cd\u7684\u5f71\u54cd":9,"\u6240\u4ee5init_hook\u5c3d\u91cf\u4f7f\u7528":24,"\u6240\u6709\u5b57\u7b26\u90fd\u5c06\u8f6c\u6362\u4e3a\u8fde\u7eed\u6574\u6570\u8868\u793a\u7684id\u4f20\u7ed9\u6a21\u578b":13,"\u6240\u6709\u6587\u4ef6\u5217\u8868":24,"\u6240\u6709\u7684":4,"\u6240\u6709\u7684\u4e0b\u8f7d\u7ebf\u7a0b\u5173\u95ed\u65f6":3,"\u6240\u6709\u914d\u7f6e\u5728":13,"\u6240\u8c13\u65f6\u95f4\u6b65\u4fe1\u606f":24,"\u624d\u4f1a\u91ca\u653e\u8be5\u6bb5\u5185\u5b58":24,"\u624d\u4f1astop":24,"\u624d\u80fd\u4fdd\u8bc1\u548c\u5355\u5c42\u5e8f\u5217\u7684\u914d\u7f6e\u4e2d":1,"\u6253\u5370\u7684\u65e5\u5fd7\u53d8\u591a":4,"\u6267\u884c":3,"\u6267\u884c\u5982\u4e0b\u547d\u4ee4\u5373\u53ef\u4ee5\u5173\u95ed\u8fd9\u4e2acontain":9,"\u6267\u884c\u65b9\u6cd5\u5982\u4e0b":9,"\u62a5\u9519":10,"\u62c6\u89e3":2,"\u62fc\u63a5\u6210\u4e00\u4e2a\u65b0\u7684\u5411\u91cf\u8868\u793a":13,"\u6307\u4ee4\u96c6":9,"\u6307\u5411\u4e00\u4e2alayer":2,"\u6307\u5b9a":2,"\u6307\u5b9a\u521d\u59cb\u5316\u6a21\u578b\u8def\u5f84":13,"\u6307\u5b9a\u663e\u5361\u6570\u91cf":14,"\u6307\u5b9a\u751f\u6210\u6570\u636e\u7684\u51fd\u6570":13,"\u6307\u5b9a\u7684\u8f93\u5165\u4e0d\u4f1a\u88ab":2,"\u6307\u5b9a\u8bad\u7ec3":13,"\u6307\u5b9abatch":13,"\u6307\u5b9aoutputs\u8f93\u51fa\u6982\u7387\u5c42":13,"\u633a":1,"\u633a\u597d":1,"\u6362":1,"\u6389\u7f16\u8bd1\u76ee\u5f55\u540e":4,"\u6392\u6210\u4e00\u5217\u7684\u591a\u4e2a\u5143\u7d20":0,"\u63a5\u4e0b\u6765\u4f7f\u7528":27,"\u63a5\u53e3\u4f7f\u7528\u591a\u7ebf\u7a0b\u8bfb\u53d6\u6570\u636e":24,"\u63a5\u53e3\u6709\u4e00\u4e2a":14,"\u63a5\u7740":1,"\u63a8\u8350":1,"\u63a8\u8350\u4f7f\u7528\u5c06\u672c\u5730\u7f51\u5361":9,"\u63a8\u8350\u4f7f\u7528\u6700\u65b0\u7248\u672c\u7684cudnn":4,"\u63a8\u8350\u6e05\u7406":4,"\u63a8\u8350\u76f4\u63a5\u653e\u7f6e\u5230\u8bad\u7ec3\u76ee\u5f55":23,"\u63a8\u8350\u8bbe\u7f6e":24,"\u63cf\u8ff0":4,"\u63cf\u8ff0\u4e86docker":3,"\u63d0\u4f9b\u6269\u5c55\u7684\u957f\u5ea6\u4fe1\u606f":0,"\u653e\u5fc3":1,"\u6548\u679c\u4e00\u81f4":24,"\u6548\u679c\u603b\u7ed3":13,"\u6559\u7a0b":13,"\u6570":2,"\u6570\u5fc5\u987b\u4e25\u683c\u76f8\u7b49":2,"\u6570\u636e":24,"\u6570\u636e\u4e2d0":14,"\u6570\u636e\u4f20\u8f93\u65e0\u9700label\u6570\u636e":13,"\u6570\u636e\u5904\u7406python\u6587\u4ef6\u540d":13,"\u6570\u636e\u5982\u4f55\u5b58\u50a8\u7b49\u7b49":24,"\u6570\u636e\u63d0\u4f9b":23,"\u6570\u636e\u6587\u4ef6\u5b58\u653e\u5728\u672c\u5730\u78c1\u76d8\u4e2d":23,"\u6570\u636e\u662f\u7ed9\u4e00\u6bb5\u82f1\u6587\u6587\u672c":24,"\u6570\u636e\u683c\u5f0f\u548c\u8be6\u7ec6\u6587\u6863\u8bf7\u53c2\u8003":13,"\u6570\u636e\u8f93\u5165":2,"\u6574\u4f53":1,"\u6574\u6d01":1,"\u6587\u4ef6":24,"\u6587\u4ef6\u4e2d":13,"\u6587\u4ef6\u4e3a":14,"\u6587\u672c\u4e2d\u7684\u5355\u8bcd\u7528\u7a7a\u683c\u5206\u9694":13,"\u6587\u672c\u4fe1\u606f\u5c31\u662f\u4e00\u4e2a\u5e8f\u5217":24,"\u6587\u672c\u5206\u7c7b\u95ee\u9898":13,"\u6587\u672c\u5377\u79ef\u5206\u4e3a\u4e09\u4e2a\u6b65\u9aa4":13,"\u6587\u672c\u751f\u6210":12,"\u6587\u6863":14,"\u65b0":1,"\u65b0\u5199layer":15,"\u65b9\u4fbf":1,"\u65b9\u4fbf\u8c03\u8bd5\u4f7f\u7528":17,"\u65b9\u4fbf\u90e8\u7f72\u5206\u53d1":17,"\u65c1\u8fb9":1,"\u65e0":1,"\u65e0\u6cd5\u76f4\u63a5\u4f7f\u7528":1,"\u65e0\u9700label\u76f8\u5173\u7684\u5c42":13,"\u65e9\u9910":1,"\u65f6":[0,14],"\u65f6\u5019":1,"\u65f6\u5e8f\u6a21\u578b\u5373\u4e3arnn\u6a21\u578b":13,"\u65f6\u5e8f\u6a21\u578b\u5747\u4f7f\u7528\u8be5\u811a\u672c":13,"\u662f":[1,4],"\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"\u662f\u4e00\u4e2a\u53cc\u5c42\u7684\u5e8f\u5217":0,"\u662f\u4e00\u4e2abatch":24,"\u662f\u4e00\u4e2apython\u7684":24,"\u662f\u4e00\u4e2aswig\u5c01\u88c5\u7684paddlepaddle\u5305":9,"\u662f\u4e00\u4e2aunbound":2,"\u662f\u4e00\u79cd\u4efb\u610f\u590d\u6742\u7684rnn\u5355\u5143":2,"\u662f\u4e0d\u662f\u5f88\u7b80\u5355\u5462":24,"\u662f\u4e2adataprovider\u662f\u4e0d\u662f\u8981\u505ashuffl":24,"\u662f\u4ec0\u4e48\u4e5f\u6ca1\u5173\u7cfb":24,"\u662f\u4ece\u8bad\u7ec3\u914d\u7f6e\u4f20\u5165\u7684dict\u5bf9\u8c61":24,"\u662f\u51e0\u4e4e\u4e0d\u5360\u5185\u5b58\u7684":24,"\u662f\u521d\u59cb\u5316\u65f6\u8c03\u7528\u7684\u51fd\u6570":24,"\u662f\u540c\u4e00\u4e2a\u5bf9\u8c61":24,"\u662f\u5426\u4ee5\u9006\u5e8f\u5904\u7406\u8f93\u5165\u5e8f\u5217":2,"\u662f\u5426\u4f7f\u7528\u53cc\u7cbe\u5ea6\u6d6e\u70b9\u6570":4,"\u662f\u5426\u4f7f\u7528\u8fd0\u884c\u65f6\u52a8\u6001\u52a0\u8f7dcuda\u52a8\u6001\u5e93":4,"\u662f\u5426\u4f7f\u7528gflags":4,"\u662f\u5426\u4f7f\u7528glog":4,"\u662f\u5426\u5185\u5d4cpython\u89e3\u91ca\u5668":4,"\u662f\u5426\u5bfb\u627e\u5230cuda\u5de5\u5177\u94fe":4,"\u662f\u5426\u5f00\u542f\u5355\u5143\u6d4b\u8bd5":4,"\u662f\u5426\u5f00\u542f\u8ba1\u65f6\u529f\u80fd\u5f00\u542f\u8ba1\u65f6\u529f\u80fd\u4f1a\u5bfc\u81f4\u8fd0\u884c\u7565\u6162":4,"\u662f\u5426\u5f00\u542fgpu\u529f\u80fd":3,"\u662f\u5426\u5f00\u542frdma\u652f\u6301":4,"\u662f\u5426\u7f16\u8bd1\u4e2d\u6587\u6587\u6863":4,"\u662f\u5426\u7f16\u8bd1\u542b\u6709avx\u6307\u4ee4\u96c6\u7684paddlepaddle\u4e8c\u8fdb\u5236":4,"\u662f\u5426\u7f16\u8bd1\u65f6\u8fdb\u884c\u4ee3\u7801\u98ce\u683c\u68c0\u67e5":4,"\u662f\u5426\u7f16\u8bd1\u82f1\u6587\u6587\u6863":4,"\u662f\u5426\u7f16\u8bd1gpu\u652f\u6301":4,"\u662f\u5426\u7f16\u8bd1python\u7684swig\u63a5\u53e3":4,"\u662f\u5728\u8fd0\u884c\u65f6\u6267\u884c\u7684":24,"\u662f\u5f85\u6269\u5c55\u7684\u6570\u636e":0,"\u662f\u60f3\u8981\u5171\u4eab\u7684\u53c2\u6570\u4f7f\u7528\u540c\u6837\u7684":14,"\u662f\u6570\u636e\u7f13\u5b58\u7684\u7b56\u7565":24,"\u662f\u6570\u636e\u8f93\u5165\u683c\u5f0f":24,"\u662f\u8bbe\u7f6e\u8fd9\u4e2adataprovider\u8fd4\u56de\u4ec0\u4e48\u6837\u7684\u6570\u636e":24,"\u662f\u8bbe\u7f6edataprovider\u5728\u5185\u5b58\u4e2d\u6682\u5b58\u7684\u6570\u636e\u6761\u6570":24,"\u662f\u8bbe\u7f6edataprovider\u5728\u5185\u5b58\u4e2d\u6700\u5c0f\u6682\u5b58\u7684\u6570\u636e\u6761\u6570":24,"\u662fdecoder\u7684\u6570\u636e\u8f93\u5165":2,"\u662fpaddlepaddle\u652f\u6301\u7684\u4e00\u79cd\u4efb\u610f\u590d\u6742\u7684rnn\u5355\u5143":2,"\u662fpaddlepaddle\u8d1f\u8d23\u63d0\u4f9b\u6570\u636e\u7684\u6a21\u5757":23,"\u662fpython\u7684\u4e00\u4e2a\u5173\u952e\u8bcd":24,"\u663e":13,"\u665a":1,"\u666e\u901a\u7528\u6237\u8bf7\u8d70\u5b89\u88c5\u6d41\u7a0b":8,"\u66f4\u591a\u5173\u4e8esparse\u8bad\u7ec3\u7684\u5185\u5bb9\u8bf7\u53c2\u8003":14,"\u66f4\u597d\u5730\u5b8c\u6210\u4e00\u4e9b\u590d\u6742\u7684\u8bed\u8a00\u7406\u89e3\u4efb\u52a1":2,"\u66f4\u65b0\u6a21\u5f0f":14,"\u66f4\u65b9\u4fbf\u7684\u8bbe\u7f6e\u65b9\u5f0f":14,"\u66f4\u8be6\u7ec6\u7528\u4f8b\u8bf7\u53c2\u8003\u6587\u6863":13,"\u66f4\u8be6\u7ec6\u7684\u4ecb\u7ecd\u8bf7\u53c2\u8003\u5404\u4e2a\u547d\u4ee4\u7684\u547d\u4ee4\u884c\u53c2\u6570\u6587\u6863":17,"\u66f4\u8be6\u7ec6\u7684\u7f51\u7edc\u914d\u7f6e":13,"\u66f4\u8fdb\u4e00\u6b65":2,"\u66ff\u6211\u4eec\u5b8c\u6210\u4e86\u539f\u59cb\u8f93\u5165\u6570\u636e\u7684\u62c6\u5206":2,"\u6700":1,"\u6700\u4f4e\u7ebf\u7a0b\u7684\u4e0b\u8f7d\u901f\u5ea6":3,"\u6700\u540e":1,"\u6700\u540e\u4e00\u4e2a":0,"\u6700\u540e\u4f7f\u7528":27,"\u6700\u7ec8\u5b9e\u73b0\u4e00\u4e2a\u5c42\u6b21\u5316\u7684\u590d\u6742rnn":2,"\u6700\u7ec8\u7684\u8f93\u51fa\u7ed3\u679c":2,"\u6708\u6e56":1,"\u6709":1,"\u6709100\u4e2a\u8bad\u7ec3\u6587\u4ef6":24,"\u6709\u4e24\u53e5":1,"\u6709\u503c\u7684\u4f4d\u7f6e\u53ea\u80fd\u53d61":24,"\u6709\u503c\u7684\u90e8\u5206\u53ef\u4ee5\u662f\u4efb\u4f55\u6d6e\u70b9\u6570":24,"\u6709\u90e8\u5206\u53c2\u6570\u662fpaddle\u81ea\u52a8\u751f\u6210\u7684":24,"\u670d\u52a1":1,"\u670d\u52a1\u5458":1,"\u672c\u6765":1,"\u672c\u8282\u6211\u4eec\u5c06\u4e13\u6ce8\u4e8e\u7f51\u7edc\u7ed3\u6784\u7684\u4ecb\u7ecd":13,"\u6765":1,"\u6765\u521d\u59cb\u5316\u53c2\u6570":14,"\u6765\u5b89\u88c5":10,"\u6765\u5bf9\u6bd4\u5206\u6790\u4e24\u8005\u8bed\u4e49\u76f8\u540c\u7684\u539f\u56e0":1,"\u6765\u5f15\u7528\u8fd9\u4e2aimag":9,"\u6765\u6307\u5b9a\u6bcf\u4e2apserver\u7684ip\u5730\u5740":14,"\u6765\u63a5\u53d7\u4e0d\u4f7f\u7528\u7684":24,"\u6765\u786e\u5b9a\u5bf9\u5e94\u5173\u7cfb":24,"\u6765\u81ea\u5b9a\u4e49\u4f20\u6570\u636e\u7684\u8fc7\u7a0b":23,"\u6765\u8bbe\u7f6e":14,"\u6765\u8bf4\u660e\u7b80\u5355\u7684pydataprovider\u5982\u4f55\u4f7f\u7528":24,"\u6765\u8fdb\u884c\u8bad\u7ec3":9,"\u6765\u914d\u7f6ecudnn\u7684\u5b89\u88c5\u8def\u5f84":4,"\u6765\u9884\u6d4b\u8fd9\u4e2a\u4e2d\u95f4\u7684\u8bcd":14,"\u676f\u5b50":1,"\u6784\u6210\u4e86\u8f93\u51fa\u53cc\u5c42\u5e8f\u5217\u7684\u7b2ci\u4e2asubseq":0,"\u6784\u9020gradientmachin":27,"\u6790\u597d\u7684\u914d\u7f6e\u521b\u5efa\u795e\u7ecf\u7f51\u7edc":27,"\u67e5\u770b\u5b89\u88c5\u540e\u7684paddl":10,"\u6807\u51c6\u5dee\u4e3a":14,"\u6807\u7b7e\u662f0":24,"\u6837\u4f8b\u6570\u636e\u4e3a":24,"\u6837\u4f8b\u6570\u636e\u5982\u4e0b":24,"\u6837\u672c":24,"\u6837\u672c\u95f4\u7528\u7a7a\u884c\u5206\u5f00":1,"\u6839\u636e\u4e0a\u4e00\u6b65\u89e3":27,"\u6839\u636e\u6a21\u578b\u914d\u7f6e\u6587\u4ef6\u4e2d":24,"\u683c\u5f0f\u5982\u4e0b":13,"\u68d2":13,"\u697c\u5c42":1,"\u6a21\u578b\u5b58\u50a8\u8def\u5f84":13,"\u6a21\u578b\u8bad\u7ec3\u4f1a\u770b\u5230\u8fd9\u6837\u7684\u65e5\u5fd7":13,"\u6a21\u578b\u914d\u7f6e":15,"\u6a2a\u5411\u5305\u62ec\u4e09\u4e2a\u7248\u672c":9,"\u6b21":1,"\u6b63\u5e38\u7684docker":9,"\u6b63\u6837\u672c":13,"\u6b64\u5904\u90fd\u4e3a2":1,"\u6bb5\u843d\u53ef\u4ee5\u770b\u4f5c\u662f\u4e00\u4e2a\u5d4c\u5957\u7684\u53cc\u5c42\u7684\u5e8f\u5217":2,"\u6bcf\u4e00\u4e2a\u4efb\u52a1\u6d41\u7a0b\u90fd\u53ef\u4ee5\u5206\u4e3a\u5982\u4e0b5\u4e2a\u57fa\u7840\u90e8\u5206":13,"\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65":1,"\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u90fd\u7528\u4e86\u4e0a\u4e00\u4e2a\u65f6\u95f4\u6b65\u7684\u8f93\u51fa\u7ed3\u679c":1,"\u6bcf\u4e00\u6761\u8bad\u7ec3\u6570\u636e\u90fd\u662f\u4e00\u4e2a\u6587\u4ef6":24,"\u6bcf\u4e00\u884c":24,"\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2a0\u5c42\u5e8f\u5217":0,"\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"\u6bcf\u4e2a\u5355\u5c42rnn":2,"\u6bcf\u4e2a\u5c42\u90fd\u6709\u4e00\u4e2a\u6216\u591a\u4e2ainput":13,"\u6bcf\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185\u7684\u8fd0\u7b97\u662f\u72ec\u7acb\u7684":2,"\u6bcf\u4e2a\u6837\u672c\u7531\u4e24\u90e8\u5206\u7ec4\u6210":1,"\u6bcf\u4e2a\u6837\u672c\u7684\u5b50\u53e5\u6570\u5206\u522b\u4e3a2":1,"\u6bcf\u4e2a\u72b6\u6001":2,"\u6bcf\u4e2agenerator\u5728\u6ca1\u6709\u8c03\u7528\u7684\u65f6\u5019":24,"\u6bcf\u4e2apass\u7684\u7b2c0\u4e2abatch\u5230\u5f53\u524dbatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747\u5206\u7c7b\u9519\u8bef\u7387":13,"\u6bcf\u4e2apass\u7684\u7b2c0\u4e2abatch\u5230\u5f53\u524dbatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747cost":13,"\u6bcf\u4e2asubseq\u7684\u6700\u540e\u4e00\u4e2a\u5143\u7d20\u5c31\u7b49\u4e8e\u5355\u5c42\u5e8f\u5217\u7684\u6700\u540e\u4e00\u4e2a\u5143\u7d20":1,"\u6bcf\u6b21\u90fd\u4f1a\u4ecepython\u7aef\u8bfb\u53d6\u6570\u636e":24,"\u6bcf\u884c\u4fdd\u5b58\u4e00\u6761\u6837\u672c":13,"\u6bcf\u9694\u591a\u5c11batch\u6253\u5370\u4e00\u6b21\u65e5\u5fd7":13,"\u6bd4\u5982":14,"\u6bd4\u5982\u8bbe\u7f6e\u4e00\u4e2a\u5168\u8fde\u63a5\u5c42\u7684\u53c2\u6570\u521d\u59cb\u5316\u65b9\u5f0f\u548cbias\u521d\u59cb\u5316\u65b9\u5f0f":14,"\u6bd4\u5982\u901a\u8fc7\u7528\u6237\u5bf9\u7535\u5b50\u5546\u52a1\u7f51\u7ad9\u8bc4\u8bba":13,"\u6bd4\u8f83\u53ef\u80fd\u7684\u547d\u4ee4\u5982\u4e0b":10,"\u6c34\u6e29":1,"\u6c49\u5ead":1,"\u6ca1":1,"\u6ca1\u6709\u4f5c\u7528":24,"\u6ca1\u6709\u5b89\u88c5":10,"\u6ca1\u6709\u8bbe\u7f6e":10,"\u6ce8\u610f":[1,3,24],"\u6cf3\u6c60":1,"\u6d41":1,"\u6d41\u7a0b\u5982\u4e0b":13,"\u6d44":1,"\u6d4b\u8bd5\u6570\u636e":13,"\u6d4b\u8bd5\u7684\u65f6\u5019\u9ed8\u8ba4\u4e0dshuffl":24,"\u6d4b\u8bd5\u811a\u672c\u5982\u4e0b":13,"\u6e29\u99a8":1,"\u6e90\u7801":13,"\u6e90\u7801\u6839\u76ee\u5f55":3,"\u6fc0\u6d3b\u51fd\u6570\u7c7b\u578b":13,"\u70ed\u60c5":1,"\u7136\u540e\u4ea4\u7ed9step\u51fd\u6570":2,"\u7136\u540e\u6267\u884c\u5982\u4e0b":10,"\u7136\u540e\u8fd0\u884c\u8fd9\u4e2acontainer\u5373\u53ef":9,"\u7248\u672c":10,"\u7279\u522b\u662f\u5728lstm\u7b49rnn\u4e2d":14,"\u751f\u6210\u5404\u4e2a\u5e73\u53f0\u7684makefil":4,"\u751f\u6210\u7684\u6570\u636e\u7f13\u5b58\u5728\u5185\u5b58\u91cc":14,"\u75280\u548c1\u8868\u793a":24,"\u7528\u4e86\u4e24\u4e2a\u6708\u4e4b\u540e\u8fd9\u4e2a\u663e\u793a\u5668\u5c4f\u5e55\u788e\u4e86":13,"\u7528\u4e8e\u4e0d\u652f\u6301avx\u6307\u4ee4\u96c6\u7684cpu":10,"\u7528\u6237\u4e5f\u53ef\u4ee5\u5728c":23,"\u7528\u6237\u4e5f\u53ef\u4ee5\u663e\u5f0f\u6307\u5b9a\u8fd4\u56de\u7684\u6570\u636e\u5bf9\u5e94\u5173\u7cfb":24,"\u7528\u6237\u53ea\u9700\u5b9a\u4e49rnn\u5728\u4e00\u4e2a\u65f6\u95f4\u6b65\u5185\u5b8c\u6210\u7684\u8ba1\u7b97":2,"\u7528\u6237\u53ef\u4ee5\u4f7f\u7528python\u7684":23,"\u7528\u6237\u53ef\u4ee5\u6839\u636e\u8bad\u7ec3log\u9009\u62e9test\u7ed3\u679c\u6700\u597d\u7684\u6a21\u578b\u6765\u9884\u6d4b":13,"\u7528\u6237\u53ef\u4ee5\u9009\u62e9\u5bf9\u5e94\u7248\u672c\u7684docker":9,"\u7528\u6237\u540d\u4e3a":9,"\u7528\u6237\u5728dataprovider\u4e2d\u9700\u8981\u5b9e\u73b0\u5982\u4f55\u8bbf\u95ee\u5176\u4e2d\u6bcf\u4e00\u4e2a\u6587\u4ef6":23,"\u7528\u6237\u5b9a\u4e49\u7684\u53c2\u6570\u4f7f\u7528args\u5728\u8bad\u7ec3\u914d\u7f6e\u4e2d\u8bbe\u7f6e":24,"\u7528\u6237\u63a5\u53e3":15,"\u7528\u6237\u9700\u8981\u5148\u5c06paddlepaddle\u5b89\u88c5\u5305\u4e0b\u8f7d\u5230\u672c\u5730":10,"\u7528\u6765\u505a\u9884\u6d4b\u548c\u7b80\u5355\u7684\u5b9a\u5236\u5316":9,"\u7528\u6b64\u65b9\u6cd5\u90e8\u5206\u865a\u62df\u673a\u53ef\u80fd\u68c0\u6d4b\u5230\u652f\u6301avx\u6307\u4ee4\u4f46\u662f\u5b9e\u9645\u8fd0\u884c\u4f1a\u6302\u6389":14,"\u7528\u8fc7\u4e00\u6b21\u7684\u65f6\u5019":24,"\u7531":2,"\u7531\u4e8e\u5916\u5c42\u6bcf\u4e2a\u65f6\u95f4\u6b65\u8fd4\u56de\u7684\u662f\u4e00\u4e2a\u5b50\u53e5":1,"\u7531\u4e8e\u5916\u5c42memory\u6ca1\u6709\u4efb\u4f55seq\u4fe1\u606f":1,"\u7531\u4e8e\u6570\u636e\u662f\u4e24\u6761":27,"\u7531\u4e8e\u8fd9\u4e2a\u5916\u5c42group\u91cc\u9762\u6ca1\u6709memori":1,"\u7531\u4e8edocker\u662f\u57fa\u4e8e\u5bb9\u5668\u7684\u8f7b\u91cf\u5316\u865a\u62df\u65b9\u6848":9,"\u7531\u4e8epaddlepaddle\u7684docker\u955c\u50cf\u5e76\u4e0d\u5305\u542b\u4efb\u4f55\u9884\u5b9a\u4e49\u7684\u8fd0\u884c\u547d\u4ee4":9,"\u7531\u4e8estep":2,"\u7531\u6613\u5230\u96be\u5c55\u793a4\u79cd\u4e0d\u540c\u7684\u7f51\u7edc\u914d\u7f6e":13,"\u7531\u8bcd\u8bed\u6784\u6210\u7684\u53e5\u5b50":0,"\u7535\u8111":1,"\u7684":[1,13],"\u7684\u4e00\u4e2a\u7b80\u5355\u8c03\u7528\u5982\u4e0b":2,"\u7684\u4efb\u4e00\u4e00\u79cd":14,"\u7684\u5185\u5b58":14,"\u7684\u540d\u5b57":24,"\u7684\u5747\u5300\u5206\u5e03":14,"\u7684\u5b89\u88c5\u6587\u6863":9,"\u7684\u5e73\u5747\u503c":0,"\u7684\u60c5\u51b5\u4e0b\u8d8a\u5927\u8d8a\u597d":24,"\u7684\u6570\u76ee\u4e00\u81f4":0,"\u7684\u6587\u6863":24,"\u7684\u65f6\u5019\u5982\u679c\u62a5\u4e00\u4e9b\u4f9d\u8d56\u672a\u627e\u5230\u7684\u9519\u8bef\u662f\u6b63\u5e38\u7684":10,"\u7684\u65f6\u95f4\u6b65\u4fe1\u606f\u6210\u6b63\u6bd4":14,"\u7684\u662f":24,"\u7684\u673a\u5668\u4e0a\u8fdb\u884c":3,"\u7684\u6838\u5fc3\u662f\u8bbe\u8ba1step\u51fd\u6570\u7684\u8ba1\u7b97\u903b\u8f91":2,"\u7684\u6bb5\u843d\u5b9a\u4e49\u4e3a\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":2,"\u7684\u72b6\u6001":2,"\u7684\u7f51\u6865\u6765\u8fdb\u884c\u7f51\u7edc\u901a\u4fe1":9,"\u7684\u8bdd":14,"\u7684\u8f93\u5165":2,"\u7684\u9519\u8bef":1,"\u7684demo\u5b66\u4e60\u5982\u4f55\u8fdb\u884c\u591a\u673a\u8bad\u7ec3":13,"\u7684docker\u53ef\u80fd\u7f3a\u4e4f":3,"\u7684matrix":27,"\u7684python\u5305\u662fpaddlepaddle\u7684\u8bad\u7ec3\u4e3b\u8981\u7a0b\u5e8f":9,"\u7684python\u5305\u6765\u505a\u914d\u7f6e\u6587\u4ef6\u89e3\u6790\u7b49\u5de5\u4f5c":9,"\u7684python\u9884\u6d4b\u8fc7\u7a0b":13,"\u76ee\u524d":2,"\u76ee\u524d\u652f\u6301\u4e24\u79cd":0,"\u76ee\u524d\u8fd8\u672a\u652f\u6301":2,"\u76ee\u5f55":13,"\u76ee\u5f55\u4e0b":3,"\u76f4\u5230\u8bad\u7ec3\u6536\u655b\u4e3a\u6b62":14,"\u76f4\u63a5\u52a0\u4e86\u4e00\u5c42group":1,"\u76f4\u63a5\u63d0\u53d6\u51fa\u795e\u7ecf\u7f51\u7edcoutput\u5c42\u7684\u8f93\u51fa\u7ed3\u679c":27,"\u76f8\u5173\u547d\u4ee4\u4e3a":9,"\u76f8\u5173\u7684\u6982":24,"\u76f8\u540c\u540d\u5b57\u7684\u53c2\u6570":14,"\u76f8\u5bf9":1,"\u76f8\u5bf9\u4e8epaddlepaddle\u7a0b\u5e8f\u8fd0\u884c\u65f6\u7684\u8def\u5f84":23,"\u76f8\u5f53":1,"\u770b\u4e0b\u9762\u7684\u89e3\u51b3\u65b9\u6848":14,"\u77e5\u9053\u5982\u4f55\u4ece":24,"\u793a":13,"\u795e\u7ecf\u7f51\u7edc\u5728\u8bad\u7ec3\u7684\u65f6\u5019":14,"\u795e\u7ecf\u7f51\u7edc\u7684\u8bad\u7ec3\u672c\u8eab\u662f\u4e00\u4e2a\u975e\u5e38\u6d88\u8017\u5185\u5b58\u548c\u663e\u5b58\u7684\u5de5\u4f5c":14,"\u79bb":1,"\u79f0\u4e4b\u4e3a\u53cc\u5c42\u5e8f\u5217\u7684\u4e00\u4e2a\u5b50\u5e8f\u5217":0,"\u7a0b\u5e8f\u6216\u8005\u81ea\u5b9a\u4e49\u4e00\u4e2a\u542b\u6709\u542f\u52a8\u811a\u672c\u7684imag":9,"\u7a97\u6237":1,"\u7aef\u81ea\u5b9a\u4e49\u4e00\u4e2a":23,"\u7aef\u8bfb\u53d6\u6570\u636e":14,"\u7b2c":1,"\u7b2c\u4e00\u4e2a\u53c2\u6570\u662f":24,"\u7b2c\u4e00\u4e2alast":1,"\u7b2c\u4e00\u4e2apass\u4f1a\u4ecepython\u7aef\u8bfb\u53d6\u6570\u636e":24,"\u7b2c\u4e00\u5929":1,"\u7b2c\u4e00\u6bb5\u6570\u636e\u4e3a\u8fd9\u5f20\u56fe\u7247\u7684label":24,"\u7b2c\u4e8c\u4e2a\u53c2\u6570\u662ffilenam":24,"\u7b2c\u4e8c\u6bb5\u6570\u636e\u4e3a\u8fd9\u4e2a\u56fe\u7247\u7684\u50cf\u7d20\u503c":24,"\u7b80\u5355\u4f18\u5316":3,"\u7b80\u5355\u7684\u4f7f\u7528":23,"\u7b80\u5355\u7684\u4f7f\u7528\u573a\u666f":23,"\u7b80\u5355\u7684\u4f7f\u7528\u6837\u4f8b\u4e3a":3,"\u7b80\u5355\u7684\u5168\u8fde\u63a5\u7f51\u7edc":14,"\u7b80\u5355\u7684\u542b\u6709ssh\u7684dockerfile\u5982\u4e0b":9,"\u7b80\u5355\u7684pydataprovider\u6837\u4f8b\u5c31\u8bf4\u660e\u5b8c\u6bd5\u4e86":24,"\u7b80\u76f4":1,"\u7b97\u6cd5":14,"\u7c7b\u522bid":13,"\u7c7b\u522bid\u7684\u6570\u636e\u7c7b\u578b":13,"\u7c7b\u578b\u53ef\u4ee5\u662fpaddlepaddle\u652f\u6301\u7684\u4efb\u610f\u8f93\u5165\u6570\u636e\u7c7b\u578b":0,"\u7c7b\u578b\u6765\u8bbe\u7f6e":24,"\u7eb5\u5411\u5305\u62ec\u56db\u4e2a\u7248\u672c":9,"\u7ec3":17,"\u7ecf\u5e38\u4f1a\u6d88\u8017\u6570\u5341g\u7684\u5185\u5b58\u548c\u6570g\u7684\u663e\u5b58":14,"\u7ed3\u4e0a\u8ff0\u7f51\u7edc\u7ed3\u6784\u5728amazon":13,"\u7ed9":1,"\u7ed9\u5b9aencoder\u8f93\u51fa\u548c\u5f53\u524d\u8bcd":2,"\u7ee7\u7eed\u8bad\u7ec3":24,"\u7ef4\u5ea6\u4e3aword":13,"\u7ef4\u5ea6\u662f\u7c7b\u522b\u4e2a\u6570":13,"\u7ef4\u5ea6\u662f\u8bcd\u5178\u5927\u5c0f":13,"\u7f13\u5b58\u6c60\u7684\u51cf\u5c0f":14,"\u7f13\u5b58\u8bad\u7ec3\u6570\u636e\u5230\u5185\u5b58":24,"\u7f16\u8bd1\u53c2\u6570\u9009\u9879\u6587\u4ef6":22,"\u7f16\u8bd1\u73af\u5883\u548c\u6e90\u4ee3\u7801":9,"\u7f16\u8bd1\u9009\u9879":4,"\u7f16\u8bd1\u9009\u9879\u4e3b\u8981\u63a8\u8350\u9ad8\u7ea7\u7528\u6237\u67e5\u770b":8,"\u7f16\u8bd1\u9009\u9879\u5217\u8868\u5982\u4e0b":4,"\u7f16\u8bd1paddlepaddle\u7684gpu\u7248\u672c\u5e76\u4e0d\u9700\u8981\u4e00\u5b9a\u5728\u5177\u6709gpu":3,"\u7f51\u7edc\u540d\u79f0":13,"\u7f51\u7edc\u914d\u7f6e":13,"\u7f6e\u8fd9\u4e9b\u53d8\u91cf":4,"\u800c":9,"\u800c\u4e09\u79cd\u5e8f\u5217\u6a21\u5f0f\u4e3a":24,"\u800c\u4e0d\u4f7f\u7528docker":9,"\u800c\u4e0d\u7528\u5173\u5fc3\u6570\u636e\u5982\u4f55\u4f20\u8f93\u7ed9paddlepaddl":24,"\u800c\u4e14\u9884\u6d4b\u7f51\u7edc\u901a\u5e38\u76f4\u63a5\u8f93\u51fa\u6700\u540e\u4e00\u5c42\u7684\u7ed3\u679c\u800c\u4e0d\u662f\u50cf\u8bad\u7ec3\u65f6\u4e00\u6837\u4ee5cost":27,"\u800c\u5728":[4,24],"\u800c\u5982\u679c\u6309\u987a\u5e8f\u8c03\u7528\u8fd9\u4e9bgenerator\u5c31\u4e0d\u4f1a\u51fa\u73b0\u8fd9\u4e2a\u95ee\u9898":24,"\u800c\u662f\u5c06\u6837\u672c\u7684\u5730\u5740\u653e\u5165\u53e6\u4e00\u4e2a\u6587\u672c":24,"\u800c\u662f\u76f4\u63a5\u4ece\u5185\u5b58\u7684\u7f13\u5b58\u91cc\u8bfb\u53d6\u6570\u636e":14,"\u800c\u6ca1\u6709\u6d4b\u8bd5\u6570\u636e":24,"\u800c\u7279\u5f81\u5373\u4e3a":24,"\u800c\u8fd9\u4e2a\u4e00\u822c\u8bf4\u660epaddlepaddle\u5df2\u7ecf\u5b89\u88c5\u5b8c\u6bd5\u4e86":10,"\u800c\u8fd9\u4e2a\u53d8\u91cf\u63a8\u8350\u5927\u4e8e\u8bad\u7ec3\u7684batch":24,"\u800c\u8fd9\u4e2acontext\u53ef\u80fd\u4f1a\u975e\u5e38":24,"\u800c\u975e\u9759\u6001\u52a0\u8f7dcuda\u52a8\u6001\u5e93":4,"\u800cexpand":1,"\u800cgpu\u7684\u9a71\u52a8\u548c\u8bbe\u5907\u5168\u90e8\u6620\u5c04\u5230\u4e86\u5bb9\u5668\u5185":9,"\u800cpaddlepaddle\u8fdb\u7a0b\u5e2e\u52a9\u7528\u6237\u505a\u4e86":24,"\u800crnn\u662f\u6700\u6d41\u884c\u7684\u9009\u62e9":2,"\u80fd\u591f\u5904\u7406\u53cc\u5c42\u5e8f\u5217":2,"\u80fd\u591f\u5bf9\u53cc\u5411\u5e8f\u5217\u8fdb\u884c\u5904\u7406\u7684\u6709":2,"\u80fd\u591f\u8bb0\u5f55\u4e0a\u4e00\u4e2asubseq":2,"\u811a\u672c":9,"\u811a\u672c\u53ef\u4ee5\u542f\u52a8paddlepaddle\u7684\u8bad\u7ec3\u8fdb\u7a0b\u548cpserv":9,"\u811a\u672c\u548c":9,"\u811a\u672c\u7c7b\u4f3c\u4e8e":9,"\u81ea\u52a8\u5b8c\u6210\u8fd9\u4e00\u8fc7\u7a0b":2,"\u81ea\u5b9a\u4e49\u4e00\u4e2adataprovid":23,"\u81f3\u5c11\u5177\u67093":9,"\u81f3\u6b64":[9,24],"\u8212\u9002":1,"\u82e5\u5e72\u4e2a\u53e5\u5b50\u6784\u6210\u4e00\u4e2a\u6bb5\u843d":0,"\u82e5\u8f93\u51fa\u662f\u5355\u5c42\u5e8f\u5217":0,"\u82e5\u8f93\u51fa\u662f\u53cc\u5c42\u5e8f\u5217":0,"\u83b7\u53d6\u5229\u7528one":13,"\u83b7\u53d6\u6bcf\u4e2a\u5355\u8bcd\u5de6\u53f3\u5404k\u4e2a\u8fd1\u90bb":13,"\u83b7\u53d6\u8be5\u6761\u6837\u672c\u7c7b\u522bid":13,"\u8868\u793a\u5c06\u5916\u5c42\u7684outer":1,"\u8868\u793a\u6574\u6570\u6807\u7b7e":24,"\u8868\u793a\u662f\u5426\u5141\u8bb8paddle\u6682\u5b58\u7565\u5fae\u591a\u4f59pool_size\u7684\u6570\u636e":24,"\u8868\u793a\u7a00\u758f\u7684\u5411\u91cf":24,"\u8868\u793a\u7a00\u758f\u7684\u96f6\u4e00\u5411\u91cf":24,"\u8868\u793a\u7a20\u5bc6\u7684\u6d6e\u70b9\u6570\u5411\u91cf":24,"\u8868\u793a\u8fc7\u4e8620\u4e2abatch":13,"\u8868\u793a\u8fc7\u4e862560\u4e2a\u6837\u672c":13,"\u8868\u793a\u8fd9\u4e2adataprovider\u662f\u8bad\u7ec3\u7528\u7684dataprovider\u6216\u8005\u6d4b\u8bd5\u7528\u7684":24,"\u8868\u793asubseq\u95f4\u4e0d\u5b58\u5728\u8054\u7cfb":1,"\u88ab\u6269\u5c55\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"\u8981\u6c42\u5355\u5c42\u5e8f\u5217\u542b\u6709\u5143\u7d20\u7684\u6570\u76ee":0,"\u8981\u751f\u6210\u7684\u76ee\u6807\u5e8f\u5217":2,"\u89c1":1,"\u89e3\u51b3\u529e\u6cd5\u662f":14,"\u89e3\u51b3\u529e\u6cd5\u662f\u5c06cuda":10,"\u89e3\u51b3\u65b9\u6cd5\u5f88\u7b80\u5355":10,"\u89e3\u6790\u8bad\u7ec3\u65f6\u7684\u914d\u7f6e\u6587\u4ef6":27,"\u89e3\u91ca":13,"\u8ba9\u795e\u7ecf\u7f51\u7edc\u53ef\u4ee5\u8fdb\u884c\u8bad\u7ec3":23,"\u8bad\u7ec3":9,"\u8bad\u7ec3\u6570\u636e\u975e\u5e38\u591a\u7684\u60c5\u51b5\u4e0b":24,"\u8bad\u7ec3\u6587\u4ef6\u5217\u8868":23,"\u8bad\u7ec3\u65f6\u6240\u9700\u8bbe\u7f6e\u7684\u4e3b\u8981\u53c2\u6570\u5982\u4e0b":13,"\u8bad\u7ec3\u7684\u65f6\u5019\u9ed8\u8ba4shuffl":24,"\u8bad\u7ec3\u811a\u672c":13,"\u8bad\u7ec3\u811a\u672c\u5728":13,"\u8bad\u7ec3\u8bed\u8a00\u6a21\u578b\u8ddd\u79bb":14,"\u8bad\u7ec3\u8f6e\u6b21":13,"\u8bb2\u6570\u636e\u53d1\u9001\u7ed9paddlepaddl":24,"\u8bb2\u89e3\u5982\u4f55\u4f7f\u7528\u53cc\u5c42rnn":1,"\u8bbe\u7f6e\u4e0b\u5217\u7f16\u8bd1\u9009\u9879\u65f6":4,"\u8bbe\u7f6e\u53c2\u6570\u7684\u540d\u5b57":14,"\u8bbe\u7f6e\u547d\u4ee4\u884c\u53c2\u6570":14,"\u8bbe\u7f6e\u6210":[14,24],"\u8bbe\u7f6e\u6210\u4e00\u4e2a\u5c0f\u4e00\u4e9b\u7684\u503c":14,"\u8bbe\u7f6e\u6210\u4e86\u5e8f\u5217":24,"\u8bbe\u7f6e\u6210true\u7684\u8bdd":24,"\u8bbe\u7f6e\u8f93\u5165\u7c7b\u578b":24,"\u8bc4\u4f30\u4ea7\u54c1\u7684\u8d28\u91cf":13,"\u8bcd\u6027\u6807\u6ce8":12,"\u8be5\u5c42\u795e\u7ecf\u5143\u4e2a\u6570":13,"\u8be5\u6570\u636e":24,"\u8be5\u6784\u5efa\u811a\u672c\u5145\u5206\u8003\u8651\u4e86\u7f51\u7edc\u4e0d\u7a33\u5b9a\u7684\u60c5\u51b5":3,"\u8be5\u6a21\u578b\u4f9d\u7136\u662f\u4f7f\u7528\u903b\u8f91\u56de\u5f52\u5206\u7c7b\u7f51\u7edc\u7684\u6846\u67b6":13,"\u8be5\u76ee\u5f55\u4e0b\u6709\u4e24\u4e2a\u6587\u4ef6":3,"\u8be5\u811a\u672c\u7684\u4f7f\u7528\u65b9\u6cd5\u662f":3,"\u8be5image\u57fa\u4e8eubuntu":3,"\u8be5image\u7684\u6784\u5efa\u5728dock":3,"\u8be6\u60c5\u8bf7\u53c2\u8003":27,"\u8be6\u7ec6\u6587\u6863\u53c2\u8003":14,"\u8be6\u7ec6\u7684\u53c2\u6570\u89e3\u91ca\u5982\u4e0b\u9762\u8868\u683c":13,"\u8be6\u7ec6\u7684\u547d\u4ee4\u884c\u53c2\u6570\u8bf7\u53c2\u8003":27,"\u8be6\u7ec6\u7684cmake\u4f7f\u7528\u65b9\u6cd5\u53ef\u4ee5\u53c2\u8003":4,"\u8be6\u7ec6\u89c1":0,"\u8bed\u4e49\u5b8c\u5168\u76f8\u540c":1,"\u8bf4\u660e":4,"\u8bf4\u660e\u547d\u4ee4\u884c\u53c2\u6570":10,"\u8bf7\u53c2\u8003":[9,24],"\u8bf7\u53c2\u8003\u4e0b\u8282refer":24,"\u8bf7\u53c2\u8003\u4e0b\u8ff0\u6587\u7ae0":23,"\u8bf7\u5b89\u88c5cuda":10,"\u8bf7\u5f53\u6210\u662f\u4e0d\u652f\u6301":14,"\u8bfb\u5165\u89e3\u6790\u8bad\u7ec3\u914d\u7f6e":27,"\u8bfb\u53d6\u6570\u636e":24,"\u8c03\u7528":4,"\u8c03\u7528\u4e00\u6b21":24,"\u8c03\u7528\u7b2c\u4e8c\u6b21\u7684\u65f6\u5019":24,"\u8d1f\u6837\u672c":13,"\u8d1f\u8d23\u591a\u673a\u8bad\u7ec3\u4e2d\u7684\u53c2\u6570\u805a\u5408\u5de5\u4f5c":17,"\u8d1f\u9762\u60c5\u7eea\u4e24\u7c7b":24,"\u8d77":1,"\u8def\u5f84\u53d8\u91cf\u4e3a":4,"\u8f83":1,"\u8f93\u5165":0,"\u8f93\u5165\u548c\u8f93\u51fa\u90fd\u662f\u5355\u5c42\u5e8f\u5217":2,"\u8f93\u5165\u548c\u8f93\u51fa\u90fd\u662f\u53cc\u5c42\u5e8f\u5217":2,"\u8f93\u5165n\u4e2a\u5355\u8bcd":13,"\u8f93\u51fa":0,"\u8f93\u51fa\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":2,"\u8f93\u51fa\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":2,"\u8f93\u51fa\u4e3an\u4e2aword":13,"\u8f93\u51fa\u5e8f\u5217\u7684\u7c7b\u578b":0,"\u8f93\u51fa\u5e8f\u5217\u7684\u8bcd\u8bed\u6570\u548c\u8f93\u5165\u5e8f\u5217\u4e00\u81f4":2,"\u8fc7\u4e86\u4e00\u4e2a\u5f88\u7b80\u5355\u7684recurr":1,"\u8fc7\u5b8c\u6240\u6709\u8bad\u7ec3\u6570\u636e\u5373\u4e3a\u4e00\u4e2apass":14,"\u8fd0\u884c":[9,10],"\u8fd0\u884c\u4f7f\u7528\u7684cudnn\u5c3d\u91cf\u662f\u540c\u4e00\u4e2a\u7248\u672c":4,"\u8fd0\u884c\u8fd9\u4e2acontain":9,"\u8fd0\u884cpaddlepaddle\u7684gpu\u7248\u672c\u4e00\u5b9a\u8981\u5728\u5177\u6709cuda\u7684\u673a\u5668\u4e0a\u8fd0\u884c":3,"\u8fd1":1,"\u8fd4\u56de0":24,"\u8fd4\u56de\u4e00\u4e2alist\u6216\u8005tupl":24,"\u8fd4\u56de\u6570\u636e\u5728paddlepaddle\u4e2d\u662f\u4ec5\u4ec5\u8fd4\u56de\u4e00\u6761\u5b8c\u6574\u7684\u8bad\u7ec3\u6837\u672c":24,"\u8fd4\u56de\u7684\u987a\u5e8f\u9700\u8981\u548c":24,"\u8fd4\u56debatch_size\u7684\u5927\u5c0f":24,"\u8fd8\u4f1a":1,"\u8fd8\u662f":1,"\u8fd8\u6709":1,"\u8fd9":[1,13,14],"\u8fd93\u4e2a\u5b50\u6b65\u9aa4\u53ef\u914d\u7f6e\u4e3a":13,"\u8fd9\u4e00\u8fc7\u7a0b\u5bf9\u7528\u6237\u662f\u5b8c\u5168\u900f\u660e\u7684":2,"\u8fd9\u4e09\u4e2alayer\u5c06\u5b83\u5148\u53d8\u6210\u4e00\u4e2a\u5143\u7d20":1,"\u8fd9\u4e24\u5c42\u4ec5\u662f\u4e3a\u4e86\u5c55\u793a\u5b83\u4eec\u7684\u7528\u6cd5":1,"\u8fd9\u4e2a":1,"\u8fd9\u4e2a\u4e5f\u662fpaddlepaddle\u6240\u80fd\u591f\u4fdd\u8bc1\u7684shuffle\u7c92\u5ea6":24,"\u8fd9\u4e2a\u4efb\u52a1\u7684\u914d\u7f6e\u4e3a":14,"\u8fd9\u4e2a\u4efb\u52a1\u7684dataprovider\u4e3a":14,"\u8fd9\u4e2a\u51fd\u6570\u4ee5\u4e00\u6761\u6570\u636e\u4e3a\u53c2\u6570":24,"\u8fd9\u4e2a\u51fd\u6570\u4f1a\u5728":24,"\u8fd9\u4e2a\u51fd\u6570\u5728\u521d\u59cb\u5316\u7684\u65f6\u5019\u4f1a\u88ab\u8c03\u7528":24,"\u8fd9\u4e2a\u51fd\u6570\u7684\u53c2\u6570\u662f":24,"\u8fd9\u4e2a\u521d\u59cb\u5316\u51fd\u6570\u5177\u6709\u5982\u4e0b\u53c2\u6570":24,"\u8fd9\u4e2a\u53c2\u6570\u5728\u8fd9\u4e2a\u6837\u4f8b\u91cc\u6ca1\u6709\u4f7f\u7528":24,"\u8fd9\u4e2a\u53c2\u6570\u88abpaddlepaddle\u8fdb\u7a0b\u4f20\u5165":24,"\u8fd9\u4e2a\u548c\u5728":24,"\u8fd9\u4e2a\u58f0\u660e\u57fa\u672c\u4e0a\u548cmnist\u7684\u6837\u4f8b\u4e00\u81f4":24,"\u8fd9\u4e2a\u5916\u5c42memori":1,"\u8fd9\u4e2a\u5b57\u5178\u53ef\u4ee5\u5728":24,"\u8fd9\u4e2a\u5bf9\u5e94\u5173\u7cfb\u53ef\u80fd\u4e0d\u6b63\u786e":24,"\u8fd9\u4e2a\u5bf9\u8c61\u548cprocess\u7684\u7b2c\u4e00\u4e2a\u53c2\u6570\u4e00\u81f4":24,"\u8fd9\u4e2a\u5de5\u5177\u7c7b\u63a5\u6536\u548cpydataprovider2\u4e00\u6837\u7684\u8f93\u5165\u6570\u636e":27,"\u8fd9\u4e2a\u5e8f\u5217\u6a21\u578b\u6bd4\u8f83\u590d\u6742":24,"\u8fd9\u4e2a\u5e8f\u5217\u7684\u6bcf\u4e2a\u5143\u7d20\u53c8\u662f\u4e00\u4e2a\u5e8f\u5217":2,"\u8fd9\u4e2a\u63a5\u53e3\u5e76\u4e0d\u7528\u6237\u53cb\u597d":27,"\u8fd9\u4e2a\u663e\u793a\u5668\u5f88\u68d2":13,"\u8fd9\u4e2a\u672c\u8eab\u4e0d\u662f\u4e00\u4e2a\u5f88":24,"\u8fd9\u4e2a\u6a21\u5757\u4e2d\u7684":24,"\u8fd9\u4e2a\u8bbe\u7f6e\u4e3a":24,"\u8fd9\u4e2a\u8f6f\u4ef6\u5305\u6587\u6863\u76f8\u5bf9\u5b8c\u5584":27,"\u8fd9\u4e2a\u8fc7\u7a0b\u5bf9\u7528\u6237\u4e5f\u662f\u900f\u660e\u7684":2,"\u8fd9\u4e2a\u95ee\u9898\u662fpydataprovider\u8bfb\u6570\u636e\u65f6\u5019\u7684\u903b\u8f91\u95ee\u9898":24,"\u8fd9\u4e2alayer\u7684\u8f93\u51fa\u4f1a\u4f5c\u4e3a\u6574\u4e2a":2,"\u8fd9\u4e5f\u4f1a\u6781\u5927\u51cf\u5c11\u6570\u636e\u8bfb\u5165\u7684\u8017\u65f6":14,"\u8fd9\u4e9b\u5185\u5b58\u5c31\u4e0d\u8003\u8651\u5982\u4f55\u7f29\u51cf\u4e86":14,"\u8fd9\u4e9b\u53c2\u6570\u5305\u62ecpaddle\u5b9a\u4e49\u7684\u53c2\u6570":24,"\u8fd9\u4e9b\u53d8":4,"\u8fd9\u4e9b\u53d8\u91cf\u53ea\u5728\u7b2c\u4e00\u6b21cmake\u7684\u65f6\u5019\u6709\u6548":4,"\u8fd9\u4e9b\u53d8\u91cf\u5747\u53ef\u4ee5\u4f7f\u7528":4,"\u8fd9\u4e9b\u5b50\u53e5\u7684\u957f\u5ea6\u5f80\u5f80\u4e0d\u7b49\u957f":1,"\u8fd9\u4e9b\u6570\u636e\u4f1a\u88ab\u7528\u6765\u66f4\u65b0\u53c2\u6570":14,"\u8fd9\u4e9b\u6570\u636e\u4f7f\u7528\u7684\u5185\u5b58\u4e3b\u8981\u548c\u4e24\u4e2a\u53c2\u6570\u6709\u5173\u7cfb":14,"\u8fd9\u4e9b\u6d41\u7a0b\u4e2d\u7684\u6570\u636e\u4e0b\u8f7d":13,"\u8fd9\u5176\u4e2d":14,"\u8fd9\u662f\u4e00\u79cd\u975e\u5e38\u7075\u6d3b\u7684\u6570\u636e\u7ec4\u7ec7\u65b9\u5f0f":0,"\u8fd9\u6837\u505a\u53ef\u4ee5\u6781\u5927\u7684\u51cf\u5c11\u5185\u5b58\u5360\u7528":14,"\u8fd9\u6837\u505a\u53ef\u4ee5\u907f\u514d\u5f88\u591a\u6b7b\u9501\u95ee\u9898":24,"\u8fd9\u79cd\u521d\u59cb\u5316\u65b9\u5f0f\u5728\u4e00\u822c\u60c5\u51b5\u4e0b\u4e0d\u4f1a\u4ea7\u751f\u5f88\u5dee\u7684\u7ed3\u679c":14,"\u8fd9\u79cd\u7c7b\u578b\u7684\u8f93\u5165\u5fc5\u987b\u901a\u8fc7":2,"\u8fd9\u884c\u7684\u4f5c\u7528\u662f\u8bbe\u7f6edataprovider\u7684\u4e00\u4e9b\u5c5e\u6027":24,"\u8fd9\u91cc":[14,24],"\u8fd9\u91cc\u4e3e\u4f8b\u7684\u6570\u636e\u662f\u82f1\u6587\u60c5\u611f\u5206\u7c7b\u7684\u6570\u636e":24,"\u8fd9\u91cc\u4ee5":13,"\u8fd9\u91cc\u4ee5mnist\u624b\u5199\u8bc6\u522b\u4e3a\u4f8b":24,"\u8fd9\u91cc\u4f7f\u7528\u7b80\u5355\u7684":14,"\u8fd9\u91cc\u53ef\u4ee5\u53c2\u8003paddle\u7684":22,"\u8fd9\u91cc\u6211\u4eec\u4f7f\u7528\u7b80\u5355\u7684\u6587\u672c\u6587\u4ef6\u8868\u793amnist\u56fe\u7247":24,"\u8fd9\u91cc\u6307\u5b9a\u8bcd\u5178":13,"\u8fd9\u91cc\u6ca1\u6709\u4ecb\u7ecd\u591a\u673a\u5206\u5e03\u5f0f\u8bad\u7ec3":13,"\u8fd9\u91cc\u7684":24,"\u8fd9\u91cc\u7684\u8f93\u5165\u7279\u5f81\u662f\u8bcdid\u7684\u5e8f\u5217":24,"\u8fd9\u91cc\u8981\u6ce8\u610f\u9884\u6d4b\u6570\u636e\u901a\u5e38":27,"\u8fd9\u91cc\u8bbe\u7f6e\u7684\u662f\u8fd4\u56de\u4e00\u4e2a":24,"\u8fd9\u91cc\u8bf4\u660e\u4e86\u8bad\u7ec3\u6570\u636e\u662f":24,"\u8fd9\u91cc\u91c7\u7528adam\u4f18\u5316\u65b9\u6cd5":13,"\u8fdb\u5165\u8be5\u6e90\u7801\u76ee\u5f55":3,"\u8fdb\u5165docker":9,"\u8fdc\u7a0b\u8bbf\u95ee":9,"\u8fde\u63a5":2,"\u8fde\u63a5\u8bf7\u53c2\u8003":13,"\u9002\u4e2d":1,"\u9009":1,"\u9009\u62e9":1,"\u9009\u62e9\u666e\u901acpu\u7248\u672c\u7684devel\u7248\u672c\u7684imag":9,"\u9009\u9879":4,"\u901a\u5e38\u505a\u6cd5\u662f\u4ece\u4e00\u4e2a\u6bd4\u8f83\u5927\u7684learning_rate\u5f00\u59cb\u8bd5":14,"\u901a\u5e38\u6765\u8bf4\u6267\u884c\u4e0bgrep":14,"\u901a\u5e38\u6839\u636e\u4efb\u52a1\u9700\u6c42\u8fdb\u884c\u4e0d\u540c\u8bbe\u7f6e":1,"\u901a\u77e5":1,"\u901a\u8fc7\u4e24\u4e2a\u5d4c\u5957\u7684":2,"\u901a\u8fc7\u591a\u7ec4\u8bed\u4e49\u76f8\u540c\u7684\u5355\u53cc\u5c42rnn\u914d\u7f6e":1,"\u901a\u8fc7\u5f15\u7528memory\u5f97\u5230\u8fd9\u4e2alayer\u4e0a\u4e00\u4e2a\u65f6\u523b\u7684\u8f93\u51fa":2,"\u901a\u8fc7\u5f15\u7528memory\u5f97\u5230\u8fd9\u4e2alayer\u4e0a\u4e00\u4e2a\u65f6\u523b\u8f93\u51fa":2,"\u901a\u8fc7\u7f16\u8bd1\u65f6\u6307\u5b9a\u8def\u5f84\u6765\u5b9e\u73b0\u5f15\u7528\u5404\u79cdbla":4,"\u901a\u8fc7data":2,"\u903b\u8f91\u56de\u5f52":13,"\u9053\u6b49":1,"\u9069":1,"\u90a3\u4e48":[2,24],"\u90a3\u4e480\u5c42\u5e8f\u5217\u5373\u4e3a\u4e00\u4e2a\u8bcd\u8bed":2,"\u90a3\u4e48\u53ef\u4ee5\u8ba4\u4e3a\u8bad\u7ec3\u4e0d\u6536\u655b":14,"\u90a3\u4e48\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d":23,"\u90a3\u4e48\u5982\u4f55\u5224\u65ad\u8bad\u7ec3\u4e0d\u6536\u655b\u5462":14,"\u90a3\u4e48\u5bf9\u5e94\u7684dataprovider\u65e2\u4e3a":24,"\u90a3\u4e48\u5e38\u6570\u8f93\u51fa\u6240\u80fd\u8fbe\u5230\u7684\u6700\u5c0fcost\u662f":14,"\u90a3\u4e48\u6211\u4eec\u53ef\u4ee5\u5224\u65ad\u4e3a\u8bad\u7ec3\u4e0d\u6536\u655b":14,"\u90a3\u4e48\u6536\u655b\u53ef\u80fd\u5f88\u6162":14,"\u90a3\u4e48\u6700\u597d\u5c06\u6570\u636e\u6587\u4ef6\u5728\u6bcf\u6b21\u8bfb\u53d6\u4e4b\u524d\u505a\u4e00\u6b21shuffl":14,"\u90a3\u4e48\u8bad\u7ec3\u6709\u53ef\u80fd\u4e0d\u6536\u655b":14,"\u90a3\u4e48\u8be5\u4f18\u5316\u7b97\u6cd5\u81f3\u5c11\u9700\u8981":14,"\u90a3\u4e48\u8fd9\u4e2a\u4e0b\u8f7d\u7ebf\u7a0b\u5c06\u4f1a\u5173\u95ed":3,"\u90a3\u4e48paddlepaddle\u4f1a\u7c97\u7565\u7684\u6839\u636elayer\u7684\u58f0\u660e\u987a\u5e8f":24,"\u90a3\u51cf\u5c11\u5b66\u4e60\u738710\u500d\u7ee7\u7eed\u8bd5\u9a8c":14,"\u90e8\u5206\u8001\u7684cpu\u578b\u53f7\u65e0\u6cd5\u652f\u6301\u8fd9\u6837\u7684\u6307\u4ee4":14,"\u90fd":1,"\u90fd\u4f20\u9012\u7ed9process\u51fd\u6570":24,"\u90fd\u662f\u5bf9layer1\u5143\u7d20\u7684\u62f7\u8d1d":0,"\u914d\u7f6e":1,"\u914d\u7f6e\u4e86":24,"\u914d\u7f6e\u53c2\u6570\u914d\u7f6e\u7ed9dataprovider\u7684":24,"\u914d\u7f6e\u6587\u4ef6":13,"\u914d\u7f6eapi":0,"\u9152\u5e97":1,"\u91c7\u7528multi":14,"\u91cc\u4f1a\u7ee7\u7eed\u5b89\u88c5":10,"\u91cc\u63d0\u4f9b\u4e86\u6570\u636e\u4e0b\u8f7d\u811a\u672c":13,"\u91cc\u9762\u8bfb\u53d6":24,"\u91cd\u65b0\u7f16\u8bd1paddlepaddl":14,"\u91cf\u4e5f\u53ef\u4ee5\u901a\u8fc7\u8c03\u7528cmake\u547d\u4ee4\u524d\u901a\u8fc7\u73af\u5883\u53d8\u91cf\u6307\u5b9a":4,"\u9488\u5bf9\u5185\u5b58\u548c\u663e\u5b58":14,"\u9488\u5bf9\u672c\u95ee\u9898":13,"\u94fe\u63a5\u4f55\u79cdblas\u7b49\u7b49":4,"\u9519\u8bef\u7387":13,"\u957f\u5ea6":14,"\u95f4\u63a5\u4f7f\u7528":1,"\u95f4\u9694":24,"\u9664\u4e86":24,"\u9664\u4e86boot":1,"\u9664\u8fc7data\u5c42":13,"\u9700\u8981\u5148\u5728\u6bcf\u4e2a\u8282\u70b9\u542f\u52a8":14,"\u9700\u8981\u53c2\u8003":9,"\u9700\u8981\u5c06\u5176parameter\u8bbe\u7f6e\u6210":14,"\u9700\u8981\u652f\u6301avx\u6307\u4ee4\u96c6\u7684cpu":9,"\u9700\u8981\u6ce8\u610f":24,"\u9700\u8981\u6ce8\u610f\u7684\u662f":[4,10],"\u9700\u8981\u9075\u5faa\u4ee5\u4e0b\u7ea6\u5b9a":2,"\u9884\u6d4b\u6570\u636e\u6307\u5b9atest":13,"\u9884\u6d4b\u7ed3\u679c\u4ee5\u6587\u672c\u7684\u5f62\u5f0f\u4fdd\u5b58\u5728":13,"\u9884\u6d4b\u811a\u672c":13,"\u9884\u6d4bid":13,"\u989d\u5916\u7684\u53c2\u6570":13,"\u9996\u5148":1,"\u9996\u5148\u5217\u4e3e\u903b\u8f91\u56de\u5f52\u7f51\u7edc":13,"\u9996\u5148\u6211\u4eec\u5c06\u8fd9\u4e2a\u6570\u636e\u6587\u4ef6":24,"\u9996\u5148\u8bf7\u53c2\u8003":13,"\u9aa43":13,"\u9ad8\u65af\u5206\u5e03":14,"\u9ed8\u8ba4\u4e00\u4e2apass\u4fdd\u5b58\u4e00\u6b21\u6a21\u578b":13,"\u9ed8\u8ba4\u4e0d\u8bbe\u7f6e":2,"\u9ed8\u8ba4\u4e3a\u7b2c\u4e00\u4e2a\u8f93\u5165":2,"\u9ed8\u8ba4\u503c":[0,4],"\u9ed8\u8ba4\u521d\u59cb\u72b6\u4e3a0":2,"\u9ed8\u8ba4\u5355\u4f4d\u662fbyte":3,"\u9ed8\u8ba4\u60c5\u51b5\u4e0b":14,"\u9ed8\u8ba4\u60c5\u51b5\u4e0b\u4e00\u6761\u6570\u636e":24,"adamax\u7b49":13,"amazon\u7535\u5b50\u4ea7\u54c1\u8bc4\u8bba\u6570\u636e":13,"api\u9884\u6d4b":13,"argument\u4f20\u5165":24,"argument\u5f62\u5f0f\u4f20\u5165":24,"async_sgd\u8fdb\u884c\u8bad\u7ec3\u65f6":14,"atlas\u5e93\u7684\u8def\u5f84":4,"batch\u4e2d\u5305\u542b":14,"batches\u8bbe\u7f6e\u6bcf\u9694\u591a\u5c11batch\u4fdd\u5b58\u4e00\u6b21\u6a21\u578b":13,"bool\u53c2\u6570":24,"case":[13,26],"cd\u5230\u542b\u6709dockerfile\u7684\u8def\u5f84\u4e2d":9,"check\u662ffalse\u7684\u8bdd":24,"class":14,"cmake\u53ef\u4ee5\u5c06cmake\u9879\u76ee\u6587\u4ef6":4,"cmake\u662f\u4e00\u4e2a\u8de8\u5e73\u53f0\u7684\u7f16\u8bd1\u811a\u672c":4,"cmake\u7684\u5b98\u65b9\u6587\u6863":4,"cmake\u7f16\u8bd1\u65f6\u4f1a\u9996\u5148\u5728\u7cfb\u7edf\u8def\u5f84":4,"container\u540e":9,"cost\u8fd8\u5927\u4e8e\u8fd9\u4e2a\u6570":14,"cpu\u7248\u672c":9,"cpuinfo\u770b\u770b\u662f\u5426\u6709\u8f93\u51fa\u5373\u53ef\u77e5\u9053\u662f\u5426\u652f\u6301":14,"cuda\u76f8\u5173\u7684driver\u548c\u8bbe\u5907\u6620\u5c04\u8fdbcontainer\u4e2d":9,"d\u547d\u4ee4\u5373\u53ef":4,"d\u547d\u4ee4\u6307\u5b9a":4,"dataprovider\u521b\u5efa\u7684\u65f6\u5019\u6267\u884c":24,"dataprovider\u53ef\u4ee5\u662f":24,"dataprovider\u63d0\u4f9b\u4e86\u4e24\u79cd\u7b80\u5355\u7684cache\u7b56\u7565":24,"dataprovider\u7684\u5177\u4f53\u7528\u6cd5\u548c\u5982\u4f55\u5b9e\u73b0\u4e00\u4e2a\u65b0\u7684dataprovid":23,"dataprovider\u7f13\u51b2\u6c60\u5185\u5b58":14,"decoder\u5faa\u73af\u5c55\u5f00\u7684\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u4f1a\u5f15\u7528\u5168\u90e8\u7ed3\u679c":2,"decoder\u63a5\u53d7\u4e24\u4e2a\u8f93\u5165":2,"decoder\u6bcf\u6b21\u9884\u6d4b\u4ea7\u751f\u4e0b\u4e00\u4e2a\u6700\u53ef\u80fd\u7684\u8bcd\u8bed":2,"decoer\u67b6\u6784":2,"devel\u548cdemo":9,"dim\u7684\u65b0\u7684\u5411\u91cf":13,"dim\u7ef4\u5ea6\u5411\u91cf":13,"docker\u662f\u4e00\u4e2a\u57fa\u4e8e\u5bb9\u5668\u7684\u8f7b\u91cf\u7ea7\u865a\u62df\u73af\u5883":9,"docker\u7684\u5b98\u65b9\u6587\u6863":9,"dockerfile\u548cbuild":3,"dockerfile\u662fdock":3,"dockerfile\u7684\u6587\u6863":9,"dockerfile\u7684\u6700\u4f73\u5b9e\u8df5":9,"driver\u6dfb\u52a0\u5230ld_library_path\u4e2d":10,"elec\u6d4b\u8bd5\u96c6":13,"embedding\u6a21\u578b\u9700\u8981\u7a0d\u5fae\u6539\u53d8\u6570\u636e\u63d0\u4f9b\u7684\u811a\u672c":13,"encoder\u548cdecoder\u53ef\u4ee5\u662f\u80fd\u591f\u5904\u7406\u5e8f\u5217\u7684\u4efb\u610f\u795e\u7ecf\u7f51\u7edc\u5355\u5143":2,"encoder\u8f93\u51fa":2,"entropy\u4f5c\u4e3acost":14,"export":[4,9,10],"f\u4ee3\u8868\u4e00\u4e2a\u6d6e\u70b9\u6570":24,"float":24,"generator\u4fbf\u4f1a\u5b58\u4e0b\u5f53\u524d\u7684\u4e0a\u4e0b\u6587":24,"generator\u7684\u4e0a\u4e0b\u6587\u4e2d\u5c3d\u91cf\u7559":24,"generator\u81f3\u5c11\u8c03\u7528\u4e24\u6b21\u624d\u4f1a\u77e5\u9053\u662f\u5426\u505c\u6b62":24,"gpu\u53cc\u7f13\u5b58":24,"gpu\u7248\u672c":9,"gpu\u7248\u672c\u4e8c\u8fdb\u5236":4,"group\u548c\u5355\u5c42\u5e8f\u5217\u7684\u51e0\u4e4e\u4e00\u6837":1,"group\u5916":1,"gru\u6a21\u578b":13,"gru\u6a21\u578b\u914d\u7f6e":13,"i\u4ee3\u8868\u4e00\u4e2a\u6574\u6570":24,"id\u4e3a0\u7684\u6982\u7387":13,"id\u4e3a1\u7684\u6982\u7387":13,"image\u6784\u5efa\u6e90\u7801\u653e\u7f6e\u5728":3,"image\u7684\u4e3b\u8981\u63cf\u8ff0\u6587\u4ef6":3,"image\u7684\u4e3b\u8981\u6784\u5efa\u6b65\u9aa4":3,"image\u7684\u6784\u5efa\u6b65\u9aa4":3,"import":[13,14,24,27],"include\u4e0b\u9700\u8981\u5305\u542bcbla":4,"include\u4e0b\u9700\u8981\u5305\u542bmkl":4,"init_hook\u53ef\u4ee5\u4f20\u5165\u4e00\u4e2a\u51fd\u6570":24,"int":[1,13,24],"key\u662fdata_layer\u7684\u540d\u5b57":24,"label\u662finteg":1,"layer1\u5fc5\u987b\u662f\u4e00\u4e2a0\u5c42\u5e8f\u5217":0,"layer1\u5fc5\u987b\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"layer2\u4e00\u81f4":0,"layer2\u53ef\u4ee5\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"layer2\u5fc5\u987b\u662f\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":0,"layer\u4e0d\u5173\u5fc3\u6570\u636e\u662f\u5426\u662f\u5e8f\u5217\u683c\u5f0f":1,"layer\u4e0d\u80fd\u94fe\u63a5\u5916\u5c42\u7684\u8fd9\u4e2amemori":1,"layer\u4f20\u7ed9\u4e0b\u4e00\u4e2a\u5b50\u53e5\u7684memori":1,"layer\u4f5c\u4e3a\u8f93\u51fa":27,"layer\u540e":1,"layer\u548caverag":1,"layer\u548cembed":1,"layer\u548clstmemori":1,"layer\u5c42":1,"layer\u62ff\u5230\u7684\u7528\u6237\u8f93\u5165":2,"layer\u6587\u6863":13,"layer\u7684\u4f7f\u7528\u793a\u4f8b\u5982\u4e0b":0,"ld_library_path\u7b49\u7b49":10,"ld_library_path\u91cc\u9762\u627e\u4e0d\u5230\u8fd9\u4e9b\u52a8\u6001":10,"lib\u4e0b\u9700\u8981\u5305\u542bcblas\u548catlas\u4e24\u4e2a\u5e93":4,"lib\u4e0b\u9700\u8981\u5305\u542bcblas\u5e93":4,"lib\u4e0b\u9700\u8981\u5305\u542bopenblas\u5e93":4,"lib\u76ee\u5f55\u4e0b\u9700\u8981\u5305\u542b":4,"list\u4e0d\u8bbe\u7f6e":23,"list\u4e2d":[23,24],"list\u4e2d\u7684\u4e00\u884c":24,"list\u4e2d\u7684\u6bcf\u4e00\u884c":24,"list\u4e3a\u7eaf\u6587\u672c\u6587\u4ef6":23,"list\u4e5f\u53ef\u4ee5\u653e\u7f6ehdfs\u6587\u4ef6\u8def\u5f84":23,"list\u5199\u5165\u90a3\u4e2a\u6587\u672c\u6587\u4ef6\u7684\u5730\u5740":24,"list\u5373\u4e3a":24,"list\u548ctest":23,"list\u5747\u4e3a\u672c\u5730\u7684\u4e24\u4e2a\u6587\u4ef6":23,"list\u6307\u5b9a\u7684\u6570\u636e":13,"list\u7684\u4f4d\u7f6e":13,"list\u82e5\u5e72\u6570\u636e\u6587\u4ef6\u8def\u5f84\u7684\u67d0\u4e00\u4e2a\u8def\u5f84":24,"lstm\u6a21\u578b\u7b49":13,"lstm\u6a21\u578b\u914d\u7f6e":13,"make\u548cmak":5,"mem\u4f5c\u4e3a\u5185\u5c42memory\u7684\u521d\u59cb\u72b6\u6001":1,"mem\u662f\u4e00\u4e2a\u5b50\u53e5\u7684\u6700\u540e\u4e00\u4e2a\u5411\u91cf":1,"memory\u4e0d\u80fd\u72ec\u7acb\u5b58\u5728":2,"memory\u53ea\u80fd\u5728":2,"memory\u6307\u5411\u4e00\u4e2alay":2,"memory\u7684\u521d\u59cb\u72b6\u6001":2,"memory\u7684\u66f4\u591a\u8ba8\u8bba\u8bf7\u53c2\u8003\u8bba\u6587":2,"memory\u7684i":2,"memory\u9ed8\u8ba4\u521d\u59cb\u5316\u4e3a0":2,"mkl\u7684\u8def\u5f84":4,"mkl_sequential\u548cmkl_intel_lp64\u4e09\u4e2a\u5e93":4,"mnist\u662f\u4e00\u4e2a\u5305\u542b\u6709":24,"movielens\u6570\u636e\u96c6":12,"movielens\u8bc4\u5206\u56de\u5f52":12,"name\u90fd\u662f":9,"no_avx\u7684":14,"osx\u6216\u8005\u662fwindows\u673a\u5668":9,"osx\u7684\u5b89\u88c5\u6587\u6863":9,"paddle\u5728\u8fdb\u884c\u8ba1\u7b97\u7684\u65f6\u5019\u4e3a\u4e86\u63d0\u5347\u8ba1\u7b97\u6027\u80fd":14,"paddle\u5b9a\u4e49\u7684\u53c2\u6570\u5305\u62ec":24,"paddle\u7684":10,"paddlepaddle\u4e2d":[0,2],"paddlepaddle\u4f7f\u7528\u5747\u503c0":14,"paddlepaddle\u4f7f\u7528\u8fd0\u884c\u65f6\u52a8\u6001\u8fde\u63a5cuda\u7684so":10,"paddlepaddle\u4fdd\u7559\u6dfb\u52a0\u53c2\u6570\u7684\u6743\u529b":24,"paddlepaddle\u53ef\u4ee5\u4f7f\u7528":4,"paddlepaddle\u53ef\u4ee5\u8bfb\u53d6python\u5199\u7684\u4f20\u8f93\u6570\u636e\u811a\u672c":13,"paddlepaddle\u5728\u8fd0\u884c\u65f6\u627e\u4e0d\u5230\u5bf9\u5e94\u7684config\u6587\u4ef6":10,"paddlepaddle\u5c06train":24,"paddlepaddle\u63a8\u8350\u4f7f\u7528docker\u8fdb\u884cpaddlepaddle\u7684\u90e8\u7f72\u548c":9,"paddlepaddle\u63d0\u4f9b\u4e86docker\u7684\u4f7f\u7528\u955c\u50cf":9,"paddlepaddle\u63d0\u4f9b\u6570\u4e2a\u9884\u7f16\u8bd1\u7684\u4e8c\u8fdb\u5236\u6765\u8fdb\u884c\u5b89\u88c5":8,"paddlepaddle\u63d0\u4f9b\u7684\u955c\u50cf\u5e76\u4e0d\u5305\u542b\u4efb\u4f55\u547d\u4ee4\u8fd0\u884c":9,"paddlepaddle\u652f\u6301\u975e\u5e38\u591a\u7684\u4f18\u5316\u7b97\u6cd5":14,"paddlepaddle\u652f\u6301sparse\u7684\u8bad\u7ec3":14,"paddlepaddle\u662f\u795e\u7ecf\u7f51\u7edc\u8bad\u7ec3\u5e73\u53f0":14,"paddlepaddle\u7684\u5185\u5b58\u5360\u7528\u4e3b\u8981\u5206\u4e3a\u5982\u4e0b\u51e0\u4e2a\u65b9\u9762":14,"paddlepaddle\u7684\u53c2\u6570\u4f7f\u7528\u540d\u5b57":14,"paddlepaddle\u7684\u6570\u636e\u5305\u62ec\u56db\u79cd\u4e3b\u8981\u7c7b\u578b":24,"paddlepaddle\u7684\u7f16\u8bd1\u9009\u9879\u53ef\u4ee5\u5728\u8c03\u7528cmake\u7684\u65f6\u5019\u8bbe\u7f6e":4,"paddlepaddle\u7684\u7f16\u8bd1\u9009\u9879\u662f\u53ef\u4ee5\u63a7\u5236paddlepaddle\u751f\u6210cpu":4,"paddlepaddle\u7684dock":3,"paddlepaddle\u7684python\u9884\u6d4b\u63a5\u53e3":26,"paddlepaddle\u7684ubuntu\u5b89\u88c5\u5305\u5206\u4e3a\u56db\u4e2a\u7248\u672c":10,"paddlepaddle\u76ee\u524d\u4f7f\u7528swig\u5bf9\u5176\u5e38\u7528\u7684\u9884\u6d4b\u63a5\u53e3\u8fdb\u884c\u4e86\u5c01\u88c5":27,"paddlepaddle\u76ee\u524d\u63d0\u4f9b\u4e24\u79cd\u53c2\u6570\u521d\u59cb\u5316\u7684\u65b9\u5f0f":14,"paddlepaddle\u76ee\u524d\u652f\u6301\u4f7f\u7528deb\u5305\u5b89\u88c5":10,"paddlepaddle\u8d1f\u8d23\u5b8c\u6210\u4fe1\u606f\u548c\u68af\u5ea6\u5728\u65f6\u95f4\u5e8f\u5217\u4e0a\u7684\u4f20\u64ad":2,"paddlepaddle\u8d1f\u8d23\u5b8c\u6210\u4fe1\u606f\u548c\u8bef\u5dee\u5728\u65f6\u95f4\u5e8f\u5217\u4e0a\u7684\u4f20\u64ad":2,"paddlepaddle\u8fd0\u884c\u65f6\u5982\u679c\u6ca1\u6709\u5bfb\u627e\u5230cuda\u7684driv":10,"paddlepaddle\u9700\u8981\u7528\u6237\u5728\u7f51\u7edc\u914d\u7f6e":23,"period\u8bbe\u7f6e\u6253\u5370\u53c2\u6570\u4fe1\u606f\u7b49":13,"process\u51fd\u6570":24,"process\u51fd\u6570\u662f\u5b9e\u73b0\u6570\u636e\u8f93\u5165\u7684\u4e3b\u51fd\u6570":24,"process\u51fd\u6570\u8c03\u7528\u591a\u6b21":24,"pserver\u4e3apaddlepaddle\u7684paramet":17,"pserver\u7684\u547d\u4ee4\u884c\u53c2\u6570":17,"pserver\u7ec4\u5408\u4f7f\u7528":17,"py\u6587\u4ef6\u7ed9\u51fa\u4e86\u5b8c\u6574\u4f8b\u5b50":13,"pydataprovider2\u4f1a\u5c3d\u91cf\u4f7f\u7528\u5185\u5b58":24,"pydataprovider2\u6587\u6863":27,"pydataprovider2\u7684\u4f7f\u7528":23,"pydataprovider\u4f7f\u7528\u7684\u662f\u5f02\u6b65\u52a0\u8f7d":14,"pydataprovider\u662fpaddlepaddle\u4f7f\u7528python\u63d0\u4f9b\u6570\u636e\u7684\u63a8\u8350\u63a5\u53e3":24,"python\u5305":9,"python\u53ef\u4ee5\u89e3\u9664\u6389\u5185\u90e8\u53d8\u91cf\u7684\u5f15\u7528":24,"python\u7684":9,"python\u7684swig\u63a5\u53e3\u53ef\u4ee5\u65b9\u4fbf\u8fdb\u884c\u9884\u6d4b\u548c\u5b9a\u5236\u5316\u8bad\u7ec3":4,"return":[1,13,24],"rnn\u603b\u662f\u5f15\u7528\u4e0a\u4e00\u65f6\u523b\u9884\u6d4b\u51fa\u7684\u8bcd\u7684\u8bcd\u5411\u91cf":2,"search\u7684\u751f\u6210":1,"seq\u53c2\u6570\u5fc5\u987b\u4e3afals":2,"seq\u540e":1,"seq\u5c42":1,"seq\u7684\u4f7f\u7528\u793a\u4f8b\u5982\u4e0b":0,"seq\u7c7b\u4f3c":0,"sequence\u7c7b\u578b":1,"server\u8fdb\u7a0b":17,"sh\u662fdocker":3,"shuffle\u8bad\u7ec3\u6570\u636e":24,"size\u53ef\u80fd\u4f1a\u5bf9\u8bad\u7ec3\u7ed3\u679c\u4ea7\u751f\u5f71\u54cd":14,"size\u672c\u8eab\u662f\u795e\u7ecf\u7f51\u7edc\u7684\u8d85\u53c2\u6570":14,"slot\u662finteg":1,"softmax\u8f93\u51fa":13,"sparse\u8bad\u7ec3\u7684\u6587\u6863":14,"sparse\u8bad\u7ec3\u9700\u8981\u8bad\u7ec3\u7279\u5f81\u662f":14,"state\u505a\u4e86\u4e00\u4e2a\u5168\u94fe\u63a5":1,"step\u4e2d":1,"step\u51fd\u6570\u4e2d\u7684memori":2,"step\u51fd\u6570\u5185\u90e8\u53ef\u4ee5\u81ea\u7531\u7ec4\u5408paddlepaddle\u652f\u6301\u7684\u5404\u79cdlay":2,"step\u7684recurr":1,"string\u7684\u683c\u5f0f\u6253\u5370\u51fa\u6765":17,"subseq\u7684\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2a0\u5c42\u5e8f\u5217":0,"swig_paddle\u63a5\u53d7\u7684\u539f\u59cb\u6570\u636e\u662fc":27,"tag\u5206\u522b\u4e3a":9,"train\u5373\u4e3apaddlepaddle\u7684\u8bad\u7ec3\u8fdb\u7a0b":17,"train\u5b8c\u6210\u5355\u673a\u591a\u663e\u5361\u591a\u7ebf\u7a0b\u7684\u8bad":17,"train\u7684\u547d\u4ee4\u884c\u53c2\u6570":17,"true":[1,14],"true\u7684memory\u65f6":1,"types\u7684\u8be6\u7ec6\u7528\u6cd5":1,"ubuntu\u7684deb\u5b89\u88c5\u5305\u7b49":8,"v2\u4e4b\u540e\u7684\u4efb\u4f55\u4e00\u4e2acudnn\u7248\u672c\u6765\u7f16\u8bd1\u8fd0\u884c":4,"value\u5373\u4e3asoftmax\u5c42\u7684\u8f93\u51fa":27,"value\u662f\u7279\u5f81\u503c":24,"value\u7c7b\u578b":1,"var":9,"vector\u8868\u793a\u7684\u6bcf\u4e2a\u5355\u8bcd":13,"version\u53ef\u4ee5\u6253\u5370\u51fapaddle\u7684\u7248\u672c\u4fe1\u606f\u548c\u7f16\u8bd1\u7684\u9009\u9879":22,"version\u53ef\u4ee5\u6253\u5370\u51fapaddlepaddle\u7684\u7248\u672c\u548c\u7f16\u8bd1\u65f6\u4fe1\u606f":17,"version\u7684\u547d\u4ee4\u884c\u53c2\u6570":17,"yield\u6587\u672c\u4fe1\u606f\u548c\u7c7b\u522bid":13,__main__:27,__name__:27,abov:24,act:[1,13,14],act_typ:13,activ:13,adadelta:[13,14],adagrad:13,adam:13,adamoptim:13,afi:24,agg_level:[0,1],aggregatelevel:[0,1],all:[2,24],allow:13,alreadi:10,also:13,append:[1,24],apt:[9,10],arg:[3,13,24],around:24,arrai:27,assert:27,atla:4,atlas_root:4,averag:1,avg:13,avgcost:13,avgpool:[0,1,13],avx:[9,14],bag:13,baidu:[9,10],batch:13,batch_siz:[1,13,14],batchsiz:1,beam:1,beam_search:2,befor:14,bias_attr:[1,14],bias_param:14,binari:13,bla:4,bool:13,boot:2,boot_lay:1,both:13,bow:13,build:[3,9],cach:[13,14,23],cache_pass_in_mem:[13,14,24],cachetyp:[13,14,24],calc_batch_s:24,call:13,can:13,can_over_batch_s:24,cat:9,categori:13,check:[1,10,24],check_fail_continu:24,chines:12,chpasswd:9,classif:13,classification_cost:[1,13,14],classification_error_evalu:13,close:24,cmake:4,cmd:9,cnn:13,code:[3,14,24,27],com:[9,10],comment:[1,13],compil:[10,22],conf:[1,27],config:[10,13],config_arg:13,config_pars:27,connect:13,contain:[13,24],context:24,context_len:13,context_start:13,convert:[13,24,27],couldn:10,cpp:[10,13],cpu:[9,10,24],cpuinfo:9,createfromconfigproto:27,cross:[13,14],cuda_so:9,cudastat:10,cudasuccess:10,cudnn:4,cudnn_root:4,cudnnv5:4,current:[13,24],currentcost:13,currentev:13,dalla:24,data:[1,10],data_config:27,data_initialz:13,data_lay:[1,13,14,24],dataprovid:[13,14],dataprovider_bow:13,dataprovider_emb:13,dataproviderconvert:27,dataset:13,deb:10,debian:10,decod:2,decor:24,def:[1,13,14,24,27],defin:[13,14,24],define_py_data_sources2:[13,24],delar:13,demo:[9,13],dense_vector:[24,27],describ:13,descript:26,detail:26,dev:9,devel:9,devic:9,devices:9,dict:[13,24],dict_dim:[1,14],dict_fil:[1,13],dictionai:13,dictionari:[13,14,24],dictrionari:13,differ:13,dim:13,dimens:[13,14],dir:13,doc:27,documentari:24,dpkg:10,driver:10,dso_handl:10,dtype:27,dump_config:17,dure:[13,24],dwith_avx:14,dynam:24,each:[13,24],each_pixel_str:24,each_sequence:[0,1],each_word:24,echo:9,either:13,els:[1,9,13],emb:[1,13,14],emb_group:1,emb_sum:14,embed:12,embedding_lay:[1,13,14],entropi:13,enumer:13,equal:1,error:[10,13],error_clipping_threshold:1,etc:9,eval:13,exampl:13,expand:[0,1],expand_a:[0,1],expand_level:[0,1],expandlevel:[0,1],expose:9,extralayerattribut:1,f0831:10,fail:[1,10],fals:[13,14],fc_layer:[1,13,14],fc_param:14,fdata:1,featur:[13,24],feature_a:14,feature_b:14,festiv:24,file:[13,24],file_list:24,file_nam:[1,13],filenam:[14,24],fill:13,find:10,first:[0,13],float32:27,fly:13,forwardtest:27,framework:13,from:[2,9,13,14,24,27],from_sequence:[0,1],from_timestep:0,full_matrix_project:1,fulli:13,func:24,gate_act:1,gdebi:10,gener:[13,24],generatedinput:2,get:[9,10,13,24],get_config_arg:13,get_data:13,get_sample_from_lin:14,github:10,give:24,given:13,globe:24,gpu:[9,10],gradient_clipping_threshold:13,gradientmachin:27,grep:9,group:1,group_input:1,gru:13,gru_siz:13,gserver:1,hassubseq:1,help:27,hidden:14,hidden_a:14,hidden_b:14,hidden_dim:1,hierach:2,hint:27,hl_cuda_devic:10,hl_dso_load:10,hook2:1,hook:1,host:9,hot:13,hous:24,howardjohnson:1,http:10,ignor:24,imag:[9,14],imagenet:12,img:24,inarg:27,includ:13,init:13,init_hook:[1,13,23],init_model_path:13,initi:[13,24],initial_max:14,initial_mean:14,initial_min:14,initial_std:14,initpaddl:27,inner:14,inner_mem:1,inner_rnn_output:1,inner_rnn_st:1,inner_step:1,input:[0,1,2,13,14,24],input_typ:[1,13,14,23],instal:5,insuffici:10,integ:[13,24],integer_sequ:[14,24],integer_valu:[1,13,14,24],integer_value_sequ:[1,13],integer_value_sub_sequ:1,invok:24,ipt:14,is_predict:13,is_train:24,isinst:27,iterat:24,job:13,join:1,kernel:9,kwarg:[1,13,24],l2regular:13,label:[1,13,14,24],label_dim:[1,13],label_list:1,lake:24,last:[0,1],later:13,latest:[3,9],layer1:0,layer2:0,layer:[0,1,2,13],ld_library_path:10,learning_method:13,learning_r:[13,14],len:[1,13,24],level:2,lib64:[9,10],lib:4,libcuda:9,libnvidia:9,librari:10,line:1,link:2,list:[13,23,24],load_data_arg:27,loadparamet:27,local:[4,10],log_period:13,logger:24,look:[13,24],loss:13,lowest_dl_speed:3,lstm:[1,13],lstm_averag:1,lstm_expand:1,lstm_group:1,lstm_group_input:1,lstm_input:1,lstm_last:1,lstm_layer_attr:1,lstm_nest_group:1,lstm_output:1,lstm_size:13,lstmemori:1,lstmemory_group:1,mac:9,machin:2,main:27,maintainer:9,make:[10,24],make_diagram:17,maxid:13,maxid_lay:13,maxpool:0,mean:[13,14],mem:1,memori:1,merge_model:17,method:24,min_pool_s:[14,24],mixed_lay:1,mkdir:9,mkl:4,mkl_core:4,mkl_root:4,mnist:24,mnist_model:27,mnist_provid:24,mnist_train:24,model_config:27,modul:[13,24],momentum:[13,14],movi:24,must:10,name:[1,9,13,14,24],necessari:13,need:13,neg:[13,24],nest:1,net:9,neural:2,next:24,no_cache:24,no_sequence:24,noavx:[9,10],none:[13,24,27],normal:9,note:10,now:2,nullptr:10,num:13,num_pass:13,nvidia:9,obj:[13,24],object:[13,24],off:[3,4,10,14,22],omit:[13,14],on_init:24,onli:[2,13],open:[1,13,14,24],openbla:4,openblas_root:4,openssh:9,opt:4,other:13,out:[1,2],outer:1,outer_mem:1,outer_rnn_st:1,outer_step:1,outlin:26,output:[1,13,14],outsid:24,paddl:[1,3,9,10,13,14,17],paddle_gpu:3,paddle_ssh:9,paddle_ssh_machin:9,paddledev:9,paddlepaddl:[9,10,22,27],param_attr:14,paramattr:14,paramet:13,parse_config:27,pass:[13,14,24],path:[10,13],period:13,permitrootlogin:9,pixel:24,pixels_float:24,pixels_str:24,place:24,pleas:10,pool:0,pool_siz:24,pooling_typ:[0,1,13,14],posit:[13,24],pred:13,predict_output_dir:13,predict_sampl:27,preprocess:13,print:27,proc:[9,14],process2:1,process:[1,13,14,24],process_pr:13,process_seq:1,process_subseq:1,properli:13,provid:1,pserver:14,pull:9,put:13,py_paddl:[9,27],pydataprovid:[14,23],pydataprovider2:[13,24,27],pydataproviderwrapp:13,python:[13,14],quick_start:13,rang:13,rank:13,rare:24,read:[13,24],read_next_from_fil:14,real_process:24,recurrent_group:[1,2],refer:23,reference_cblas_root:4,reffer:4,regular:13,releas:10,repres:13,represent:13,resnet:12,result:[13,24],revers:2,rmsprop:13,rnn:2,rnn_data_provid:1,rnn_state:1,roce:9,root:9,run:9,runtim:[10,24],same:[13,24],sampl:[13,24],save:[13,24],save_dir:13,saw:24,sbin:9,script:3,second:13,sed:9,see:13,sentanc:14,sentenc:24,sentiment:24,sentimental_provid:24,separ:13,seq:[0,1],seq_pool:0,seq_typ:24,seqlastin:1,sequel:24,sequenc:[1,2],sequence:24,sequence_conv_pool:13,sequence_layer_group:1,sequence_nest_layer_group:1,sequence_nest_rnn:1,sequence_nest_rnn_readonly_memori:1,sequence_rnn:1,sequencegen:1,sequencestartposit:1,sequencetyp:24,server:9,set:[1,13,14,24],setup:13,should:2,should_shuffl:24,shuf:14,shuffl:14,sigmoidactiv:1,simple_gru:13,simple_lstm:13,size:[1,13,14,24],softmax:[13,14],softmax_param:14,softmaxactiv:[1,13,14],sourc:13,spars:[13,14],sparse_binary_vector:[13,14,24],sparse_float_vector:24,sparse_upd:14,sparse_vector:14,specifi:[10,13],split:[1,13,24],src_root:27,ssh:9,sshd:9,sshd_config:9,stat:13,state:2,state_act:1,staticinput:2,step:[1,2],stop:9,store:13,string:24,strip:[1,13],structur:13,stun:24,sub:1,sub_sequence:24,subseq:[0,2],subsequenceinput:1,sudo:10,sumpool:14,support:9,sure:10,swig_paddl:27,system:14,tag:3,take:24,tanhactiv:1,tbd:[1,25],team:9,test:[1,13,23],test_data:27,test_list:[13,24],test_recurrentgradientmachin:1,text:[13,24],text_conv:13,them:13,thi:[13,24],thing:24,timestep:0,tmp:24,tour_train_wdseg:1,train:10,train_list:[13,24],trainer:[13,24,27],trainer_config:[13,23,24,27],trainer_config_help:[13,14,24],trainer_count:14,trainerintern:13,trainermain:10,travel:24,trn:13,tst:13,turn:2,two:13,txt:[13,24],type:[13,24],unk_idx:13,updat:9,use:[13,26],use_dynamic_ord:24,use_gpu:[13,14,27],usepam:9,user:13,usr:[4,9,10],valid:10,valu:[1,13,24,27],version:[9,10],via:10,want:24,what:13,when:24,which:13,whole:24,wilder:24,window:9,with_avx:[4,10,22],with_doc:4,with_doc_cn:4,with_doubl:[10,22],with_double:4,with_dso:4,with_gflag:[10,22],with_gflags:4,with_glog:[4,10,22],with_gpu:[3,4,10,22],with_metric_learn:[10,22],with_predict_sdk:[10,22],with_python:[4,10,22],with_rdma:[4,10,22],with_style_check:4,with_swig_py:4,with_testing:4,with_tim:[10,22],with_timer:4,without:9,wonder:24,word2vec:14,word:[1,2,12],word_dict:[1,13],word_dim:[1,13],word_id:[14,24],word_slot:1,word_slot_list:1,word_vector:13,xarg:9,yield:[1,13,14,24],you:[10,24],your_host_machine:9,your_param_name:14},titles:["\u652f\u6301\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165\u7684Layer","\u53cc\u5c42RNN\u914d\u7f6e\u4e0e\u793a\u4f8b","Recurrent Group\u6559\u7a0b","\u6784\u5efaPaddlePaddle Docker Image","\u8bbe\u7f6ePaddlePaddle\u7684\u7f16\u8bd1\u9009\u9879","\u4f7f\u7528cmake\u7f16\u8bd1PaddlePaddle","\u5b89\u88c5\u7f16\u8bd1PaddlePaddle\u9700\u8981\u7684\u4f9d\u8d56","make\u548cmake install","\u7f16\u8bd1\u4e0e\u5b89\u88c5","\u5b89\u88c5PaddlePaddle\u7684Docker\u955c\u50cf","\u4f7f\u7528deb\u5305\u5728Ubuntu\u4e0a\u5b89\u88c5PaddlePaddle","\u96c6\u7fa4\u8bad\u7ec3","\u4f7f\u7528\u793a\u4f8b","PaddlePaddle\u5feb\u901f\u5165\u95e8\u6559\u7a0b","PaddlePaddle\u5e38\u89c1\u95ee\u9898","PaddlePaddle\u6587\u6863","&lt;no title&gt;","\u547d\u4ee4\u884c\u53c2\u6570","&lt;no title&gt;","&lt;no title&gt;","paddle pserver\u7684\u547d\u4ee4\u884c\u53c2\u6570","paddle train\u7684\u547d\u4ee4\u884c\u53c2\u6570","paddle version\u7684\u547d\u4ee4\u884c\u53c2\u6570","PaddlePaddle\u7684\u6570\u636e\u63d0\u4f9b(DataProvider)\u4ecb\u7ecd","PyDataProvider2\u7684\u4f7f\u7528","\u81ea\u5b9a\u4e49\u4e00\u4e2aDataProvider","\u7528\u6237\u63a5\u53e3","PaddlePaddle\u7684Python\u9884\u6d4b\u63a5\u53e3"],titleterms:{"\u4e0b\u8f7d\u548c\u8fd0\u884cdocker\u955c\u50cf":9,"\u4ecb\u7ecd":23,"\u4f18\u5316\u7b97\u6cd5":13,"\u4f7f\u7528\u6307\u5357":15,"\u4f7f\u7528\u6982\u8ff0":13,"\u4f7f\u7528\u793a\u4f8b":12,"\u4f7f\u7528\u811a\u672c\u6784\u5efapaddlepaddl":3,"\u4f7f\u7528cmake\u7f16\u8bd1paddlepaddl":5,"\u4f7f\u7528deb\u5305\u5728ubuntu\u4e0a\u5b89\u88c5paddlepaddl":10,"\u5185\u5b58\u4e0d\u591f\u7528\u7684\u60c5\u51b5":24,"\u51cf\u5c11\u6570\u636e\u8f7d\u5165\u7684\u8017\u65f6":14,"\u51cf\u5c11dataprovider\u7f13\u51b2\u6c60\u5185\u5b58":14,"\u5229\u7528\u66f4\u591a\u7684\u8ba1\u7b97\u8d44\u6e90":14,"\u52a0\u901f\u8bad\u7ec3\u901f\u5ea6":14,"\u5377\u79ef\u6a21\u578b":13,"\u53c2\u6570\u5185\u5b58":14,"\u53c2\u8003":24,"\u53cc\u5c42rnn\u4ecb\u7ecd":2,"\u53cc\u5c42rnn\u7684\u4f7f\u7528":2,"\u53cc\u5c42rnn\u914d\u7f6e\u4e0e\u793a\u4f8b":1,"\u53cc\u8fdb\u53cc\u51fa":1,"\u53ef\u80fd\u7684\u5185\u5b58\u6cc4\u9732\u95ee\u9898":24,"\u53ef\u80fd\u9047\u5230\u7684\u95ee\u9898":10,"\u547d\u4ee4\u884c\u53c2\u6570":[13,17,26],"\u548c":0,"\u56fe\u50cf":12,"\u57fa\u672c\u539f\u7406":2,"\u5982\u4f55\u5171\u4eab\u53c2\u6570":14,"\u5982\u4f55\u51cf\u5c11paddlepaddle\u7684\u5185\u5b58\u5360\u7528":14,"\u5982\u4f55\u521d\u59cb\u5316\u53c2\u6570":14,"\u5982\u4f55\u52a0\u901fpaddlepaddle\u7684\u8bad\u7ec3\u901f\u5ea6":14,"\u5982\u4f55\u9009\u62e9sgd\u7b97\u6cd5\u7684\u5b66\u4e60\u7387":14,"\u5b89\u88c5":[8,13],"\u5b89\u88c5\u7f16\u8bd1paddlepaddle\u9700\u8981\u7684\u4f9d\u8d56":6,"\u5b89\u88c5paddlepaddle\u7684docker\u955c\u50cf":9,"\u5e38\u7528\u6a21\u578b":12,"\u5e38\u89c1\u95ee\u9898":15,"\u5e8f\u5217\u6a21\u578b\u6570\u636e\u63d0\u4f9b":24,"\u5f00\u53d1\u6307\u5357":15,"\u6027\u80fd\u95ee\u9898":9,"\u603b\u4f53\u6548\u679c\u603b\u7ed3":13,"\u6216\u8005\u662f":14,"\u63a8\u8350":12,"\u652f\u6301\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165\u7684layer":0,"\u6570\u636e\u5411\u6a21\u578b\u4f20\u9001":13,"\u6570\u636e\u63d0\u4f9b":26,"\u6570\u636e\u683c\u5f0f\u51c6\u5907":13,"\u65f6\u5e8f\u6a21\u578b":13,"\u6784\u5efapaddlepaddl":3,"\u6982\u8ff0":[0,2],"\u6a21\u578b\u4e2d\u7684\u914d\u7f6e":1,"\u6ce8\u610f\u4e8b\u9879":[9,24],"\u751f\u6210\u6d41\u7a0b\u7684\u4f7f\u7528\u65b9\u6cd5":2,"\u7528\u6237\u63a5\u53e3":26,"\u76f8\u5173\u6982\u5ff5":2,"\u793a\u4f8b1":1,"\u793a\u4f8b2":1,"\u793a\u4f8b3":1,"\u793a\u4f8b4":1,"\u795e\u7ecf\u5143\u6fc0\u6d3b\u5185\u5b58":14,"\u7b80\u5355\u7684\u4f7f\u7528\u573a\u666f":24,"\u7b97\u6cd5\u6559\u7a0b":15,"\u7f16\u8bd1":8,"\u7f16\u8bd1\u4e0e\u5b89\u88c5":8,"\u7f51\u7edc\u7ed3\u6784":13,"\u81ea\u5b9a\u4e49\u4e00\u4e2adataprovid":25,"\u81ea\u7136\u8bed\u8a00\u5904\u7406":12,"\u8bad\u7ec3\u6a21\u578b":13,"\u8bad\u7ec3\u6d41\u7a0b\u7684\u4f7f\u7528\u65b9\u6cd5":2,"\u8bbe\u7f6epaddlepaddle\u7684\u7f16\u8bd1\u9009\u9879":4,"\u8bcd\u5411\u91cf\u6a21\u578b":13,"\u8bfb\u53d6\u53cc\u5c42\u5e8f\u5217\u7684\u65b9\u6cd5":1,"\u8f93\u5165":2,"\u8f93\u5165\u4e0d\u7b49\u957f":1,"\u8f93\u5165\u793a\u4f8b":2,"\u8f93\u51fa":2,"\u8f93\u51fa\u65e5\u5fd7":13,"\u8fdc\u7a0b\u8bbf\u95ee\u95ee\u9898\u548c\u4e8c\u6b21\u5f00\u53d1":9,"\u903b\u8f91\u56de\u5f52\u6a21\u578b":13,"\u9047\u5230":14,"\u914d\u7f6e\u4e2d\u7684\u6570\u636e\u52a0\u8f7d\u5b9a\u4e49":13,"\u9644\u5f55":13,"\u96c6\u7fa4\u8bad\u7ec3":11,"\u975e\u6cd5\u6307\u4ee4":14,"\u9884\u6d4b":[13,26],"beam_search\u7684\u751f\u6210":1,"blas\u76f8\u5173\u7684\u7f16\u8bd1\u9009\u9879":4,"bool\u578b\u7684\u7f16\u8bd1\u9009\u9879":4,"config\u6587\u4ef6\u627e\u4e0d\u5230":10,"cudnn\u76f8\u5173\u7684\u7f16\u8bd1\u9009\u9879":4,"driver\u627e\u4e0d\u5230":10,"group\u6559\u7a0b":2,"make\u548cmak":7,"paddlepaddle\u5e38\u89c1\u95ee\u9898":14,"paddlepaddle\u5feb\u901f\u5165\u95e8\u6559\u7a0b":13,"paddlepaddle\u63d0\u4f9b\u7684docker\u955c\u50cf\u7248\u672c":9,"paddlepaddle\u6587\u6863":15,"paddlepaddle\u7684\u6570\u636e\u63d0\u4f9b":23,"paddlepaddle\u7684\u7f16\u8bd1\u9009\u9879":4,"paddlepaddle\u7684bool\u578b\u7f16\u8bd1\u9009\u9879":4,"paddlepaddle\u7684cblas\u7f16\u8bd1\u9009\u9879":4,"paddlepaddle\u7684python\u9884\u6d4b\u63a5\u53e3":27,"pserver\u7684\u547d\u4ee4\u884c\u53c2\u6570":20,"pydataprovider2\u7684\u4f7f\u7528":24,"python\u6570\u636e\u52a0\u8f7d\u811a\u672c":13,"so\u627e\u4e0d\u5230":10,"subseq\u95f4\u65e0memori":1,"subseq\u95f4\u6709memori":1,"train\u7684\u547d\u4ee4\u884c\u53c2\u6570":21,"version\u7684\u547d\u4ee4\u884c\u53c2\u6570":22,algorithm:13,appendix:13,architectur:13,argument:13,cach:24,command:13,configur:13,content:14,convolut:13,cuda:[4,10],data:13,dataprovid:23,docker:3,expand_lay:0,first_seq:0,illeg:14,image:3,init_hook:24,input_typ:24,instal:7,install:13,instruct:14,last_seq:0,libcudart:10,libcudnn:10,line:13,log:13,logist:13,memori:2,model:13,network:13,optimiz:13,overview:13,paddl:[20,21,22],pooling_lay:0,predict:13,prepar:13,provid:[13,24],recurr:2,refer:24,regress:13,script:13,sequenc:13,summari:13,time:13,train:13,transfer:13,vector:13,word:13}})
\ No newline at end of file
Search.setIndex({envversion:49,filenames:["algorithm/rnn/hierarchical-layer","algorithm/rnn/hierarchical-rnn","algorithm/rnn/rnn-tutorial","build/docker/build_docker_image","build_and_install/cmake/compile_options","build_and_install/cmake/index","build_and_install/cmake/install_deps","build_and_install/cmake/make_and_install","build_and_install/index","build_and_install/install/docker_install","build_and_install/install/ubuntu_install","cluster/index","demo/index","demo/quick_start/index","index","ui/cmd/dump_config","ui/cmd/index","ui/cmd/make_diagram","ui/cmd/merge_model","ui/cmd/paddle_pserver","ui/cmd/paddle_train","ui/cmd/paddle_version","ui/data_provider/index","ui/data_provider/pydataprovider2","ui/data_provider/write_new_dataprovider","ui/index","ui/predict/swig_py_paddle"],objects:{},objnames:{},objtypes:{},terms:{"0000x":13,"000\u5f20\u7070\u5ea6\u56fe\u7247\u7684\u6570\u5b57\u5206\u7c7b\u6570\u636e\u96c6":23,"00186201e":26,"04\u4e2d\u6b63\u786e":10,"08823112e":26,"0\u5c42\u5e8f\u5217":0,"0b1":10,"10\u4ee5\u4e0a\u7684linux":9,"10\u7ef4\u7684\u6574\u6570\u503c":23,"10gbe":9,"10m":3,"12194102e":26,"12\u7248\u672c\u6d4b\u8bd5\u901a\u8fc7":3,"12\u7248\u672c\u7684\u60c5\u51b5\u4e0b\u5e76\u6ca1\u6709\u6d4b\u8bd5":3,"15501715e":26,"15mb":13,"16mb":13,"1\u7684\u8bdd":23,"252kb":13,"25639710e":26,"25k":13,"27787406e":26,"28\u7684\u50cf\u7d20\u7070\u5ea6\u503c":23,"28\u7684\u7a20\u5bc6\u5411\u91cf\u548c\u4e00\u4e2a":23,"2\u4e2a\u5b50\u53e5":1,"2\u53e5\u53cc\u5c42\u5e8f\u5217\u548c5\u53e5\u5355\u5c42\u5e8f\u5217\u7684\u6570\u636e\u5b8c\u5168\u4e00\u6837":1,"2\u8868\u793a\u4e00\u6b21\u8fc72\u53e5\u53cc\u5c42\u5e8f\u5217":1,"2\u8fdb\u884c\u8fdb\u4e00\u6b65\u6f14\u5316":13,"32777140e":26,"36540484e":26,"3\u4e2a\u5b50\u53e5":1,"40gbe":9,"43630644e":26,"48565123e":26,"48684503e":26,"49316648e":26,"50k":3,"51111044e":26,"53018653e":26,"56gbe":9,"5\u5230\u672c\u5730\u73af\u5883\u4e2d":10,"5\u8868\u793a\u4e00\u6b21\u8fc75\u53e5\u5355\u5c42\u5e8f\u5217":1,"70634608e":26,"72296313e":26,"85625684e":26,"93137714e":26,"96644767e":26,"99982715e":26,"9\u7684\u6570\u5b57":23,"\u4e00":1,"\u4e00\u4e2a0\u5c42\u5e8f\u5217":0,"\u4e00\u4e2a0\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u6269\u5c55\u6210\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u6269\u5c55\u6210\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":0,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u8fdb\u5165":2,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":0,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u6216\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":0,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u8fdb\u5165":2,"\u4e00\u4e2a\u53cc\u5c42rnn\u7531\u591a\u4e2a\u5355\u5c42rnn\u7ec4\u6210":2,"\u4e00\u4e2a\u53ef\u8c03\u7528\u7684\u51fd\u6570":2,"\u4e00\u4e2a\u6587\u4ef6":23,"\u4e00\u4e2a\u72ec\u7acb\u7684\u5143\u7d20":0,"\u4e00\u4e2a\u72ec\u7acb\u7684\u8bcd\u8bed":0,"\u4e00\u4e2alabel":1,"\u4e00\u4e2alogging\u5bf9\u8c61":23,"\u4e00\u4e2apass\u8868\u793a\u8fc7\u4e00\u904d\u6240\u6709\u8bad\u7ec3\u6837\u672c":13,"\u4e00\u4eba":1,"\u4e00\u5171\u670910\u4e2a\u6837\u672c":1,"\u4e00\u5171\u67094\u4e2a\u6837\u672c":1,"\u4e00\u53e5\u8bdd\u662f\u7531\u8bcd\u8bed\u6784\u6210\u7684\u5e8f\u5217":2,"\u4e00\u65e9":1,"\u4e00\u6761":23,"\u4e00\u6b21\u6027\u676f\u5b50":1,"\u4e00\u81f4":1,"\u4e00\u81f4\u7684\u7279\u5f81":23,"\u4e00\u822c\u60c5\u51b5\u4e0b":22,"\u4e00\u822c\u63a8\u8350\u8bbe\u7f6e\u6210true":23,"\u4e00\u884c\u4e3a\u4e00\u4e2a\u6837\u672c":13,"\u4e00\u884c\u5bf9\u5e94\u4e00\u4e2a\u6570\u636e\u6587\u4ef6":22,"\u4e0a\u7684\u6548\u679c":13,"\u4e0a\u7f51":1,"\u4e0b\u6587\u4ee5nlp\u4efb\u52a1\u4e3a\u4f8b":2,"\u4e0b\u6b21":1,"\u4e0b\u8f7d\u8fdb\u7a0b\u4f1a\u91cd\u542f":3,"\u4e0b\u8ff0\u5185\u5bb9\u5c06\u5206\u4e3a\u5982\u4e0b\u51e0\u4e2a\u7c7b\u522b\u63cf\u8ff0":9,"\u4e0b\u975e\u5e38\u5c11\u7684\u53d8\u91cf\u5f15\u7528":23,"\u4e0b\u9762\u8fd9\u4e9blayer\u80fd\u591f\u63a5\u53d7\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165":0,"\u4e0b\u9762dataprovid":13,"\u4e0d":1,"\u4e0d\u4e00\u5b9a\u548c\u65f6\u95f4\u6709\u5173\u7cfb":23,"\u4e0d\u4f1a\u6267\u884c\u6d4b\u8bd5\u64cd\u4f5c":22,"\u4e0d\u5305\u542blabel":26,"\u4e0d\u540c\u7684\u6570\u636e\u7c7b\u578b\u548c\u5e8f\u5217\u6a21\u5f0f\u8fd4\u56de\u7684\u683c\u5f0f\u4e0d\u540c":23,"\u4e0d\u540c\u8f93\u5165\u542b\u6709\u7684\u5b50\u53e5":2,"\u4e0d\u540c\u8f93\u5165\u5e8f\u5217\u542b\u6709\u7684\u8bcd\u8bed\u6570\u5fc5\u987b\u4e25\u683c\u76f8\u7b49":2,"\u4e0d\u5c11":1,"\u4e0d\u5e94\u8be5\u88ab\u62c6\u89e3":2,"\u4e0d\u6307\u5b9a\u65f6":2,"\u4e0d\u652f\u6301avx\u6307\u4ee4\u96c6\u7684cpu\u4e5f\u53ef\u4ee5\u8fd0\u884c":9,"\u4e0d\u7f13\u5b58\u4efb\u4f55\u6570\u636e":23,"\u4e0d\u8fc7":1,"\u4e0d\u8fdc":1,"\u4e0d\u9519":1,"\u4e0d\u9700\u8981avx\u6307\u4ee4\u96c6\u7684cpu\u4e5f\u53ef\u4ee5\u8fd0\u884c":9,"\u4e0e\u8bad\u7ec3\u7f51\u7edc\u914d\u7f6e\u4e0d\u540c\u7684\u662f":13,"\u4e14":1,"\u4e14\u5e8f\u5217\u7684\u6bcf\u4e00\u4e2a\u5143\u7d20\u8fd8\u662f\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":23,"\u4e24":1,"\u4e24\u4e2a\u5d4c\u5957\u7684":2,"\u4e24\u4e2a\u6587\u6863":9,"\u4e24\u7c7b":13,"\u4e25\u91cd\u7684\u95ee\u9898":23,"\u4e2a":13,"\u4e2ayield":23,"\u4e2d":13,"\u4e2d\u5b9a\u4e49\u4f7f\u7528\u54ea\u79cddataprovider\u53ca\u5176\u53c2\u6570":22,"\u4e2d\u5b9a\u4e49\u548c\u4f7f\u7528":2,"\u4e2d\u5b9a\u4e49\u7684\u987a\u5e8f\u4e00\u81f4":23,"\u4e2d\u5bfb\u627e\u8fd9\u4e9bblas\u7684\u5b9e\u73b0":4,"\u4e2d\u7684":23,"\u4e2d\u7684\u4e8c\u8fdb\u5236\u4f7f\u7528\u4e86":9,"\u4e2d\u7684set":23,"\u4e2d\u914d\u7f6e":23,"\u4e3a":23,"\u4e3a\u4e86\u63cf\u8ff0\u65b9\u4fbf":2,"\u4e3a\u4e86\u8fd0\u884cpaddlepaddle\u7684docker\u955c\u50cf":9,"\u4e3a\u4f8b\u8fdb\u884c\u9884\u6d4b":13,"\u4e3b\u8981\u51fd\u6570\u662fprocess\u51fd\u6570":23,"\u4e3b\u8981\u5206\u4e3a\u4ee5\u4e0b\u51e0\u4e2a\u6b65\u9aa4":26,"\u4e3b\u8981\u5305\u62ec\u4e24\u90e8\u5206":13,"\u4e3b\u8981\u539f\u56e0":1,"\u4e3b\u8981\u662f\u589e\u52a0\u4e86\u521d\u59cb\u5316\u673a\u5236":23,"\u4e3b\u8981\u6b65\u9aa4\u4e3a":26,"\u4e3b\u8981\u7531\u4e8e\u65e7\u7248\u672c":3,"\u4e3b\u8981\u7684\u8f6f\u4ef6\u5305\u4e3apy_paddl":26,"\u4e4b\u95f4\u7684\u8fd0\u7b97\u662f\u72ec\u7acb\u7684":2,"\u4e5f":1,"\u4e5f\u4e0d\u5b58\u5728\u4e00\u4e2asubseq\u76f4\u63a5\u751f\u6210\u4e0b\u4e00\u4e2asubseq\u7684\u60c5\u51b5":2,"\u4e5f\u4f1a\u6254\u5230\u8fd9\u6761\u6570\u636e":23,"\u4e5f\u4f1a\u8bfb\u53d6\u76f8\u5173\u8def\u5f84\u53d8\u91cf\u6765\u8fdb\u884c\u641c\u7d22":4,"\u4e5f\u53ef\u4ee5":23,"\u4e5f\u53ef\u4ee5\u4e3ainteg":1,"\u4e5f\u53ef\u4ee5\u4f7f\u7528":23,"\u4e5f\u53ef\u4ee5\u548cpaddl":16,"\u4e5f\u53ef\u4ee5\u662f\u4e00\u4e2a\u8bcd\u8bed":2,"\u4e5f\u53ef\u4ee5\u76f4\u63a5\u6267\u884c":9,"\u4e5f\u53ef\u4ee5\u901a\u8fc7\u5982\u4e0b\u65b9\u5f0f\u9884\u6d4b":13,"\u4e5f\u53ef\u4ee5\u901a\u8fc7save":13,"\u4e5f\u53ef\u4ee5\u9884\u6d4b\u6ca1\u6709label\u7684\u6d4b\u8bd5\u96c6":13,"\u4e5f\u5c31\u662f\u5c06\u8bcd\u5411\u91cf\u6a21\u578b\u989d\u6b65":13,"\u4e5f\u5c31\u662f\u76f4\u63a5\u5199\u5185\u5b58\u7684float\u6570\u7ec4":26,"\u4e5f\u662fdecoder\u5faa\u73af\u5c55\u5f00\u7684\u4f9d\u636e":2,"\u4e5f\u9700\u8981\u4e24\u6b21\u968f\u673a\u9009\u62e9\u5230\u540c\u6837\u7684generator\u7684\u65f6\u5019":23,"\u4e7e":1,"\u4e86":1,"\u4e86\u975e\u5e38\u65b9\u4fbf\u7684\u4e8c\u8fdb\u5236\u5206\u53d1\u624b\u6bb5":9,"\u4e8c\u6b21\u5f00\u53d1\u53ef\u4ee5":9,"\u4e94\u661f\u7ea7":1,"\u4ea4\u901a":1,"\u4ea4\u901a\u4fbf\u5229":1,"\u4eba\u5458\u7b49\u7b49":3,"\u4ec5\u4ec5\u9700\u8981":23,"\u4ecb\u7ecdpaddlepaddle\u4f7f\u7528\u6d41\u7a0b\u548c\u5e38\u7528\u7684\u7f51\u7edc\u57fa\u7840\u5355\u5143\u7684\u914d\u7f6e\u65b9\u6cd5":13,"\u4ece\u4e00\u4e2aword\u751f\u6210\u4e0b\u4e00\u4e2aword":2,"\u4ece\u6587\u4ef6\u4e2d\u8bfb\u53d6\u6bcf\u4e00\u6761\u6570\u636e":23,"\u4ece\u6587\u672c\u6587\u4ef6\u4e2d\u8bfb\u53d6":23,"\u4ece\u800c\u4e0d\u80fd\u5728\u8fd0\u884c\u7f16\u8bd1\u547d\u4ee4\u7684\u65f6\u5019\u63a5\u53d7\u53c2\u6570":3,"\u4ece\u800c\u751f\u6210\u591a\u4e2agener":23,"\u4ece\u800c\u9632\u6b62\u8fc7\u62df\u5408":22,"\u4ece\u8bed\u4e49\u4e0a\u770b":2,"\u4ece\u8f93\u5165\u6570\u636e\u4e0a\u770b":1,"\u4ed6\u4eec\u662f":[9,10,16,23],"\u4ed6\u4eec\u7684imag":9,"\u4ed6\u53ef\u4ee5\u5c06\u67d0\u4e00\u4e2a\u51fd\u6570\u6807\u8bb0\u6210\u4e00\u4e2apydataprovid":23,"\u4ee3\u8868\u4e00\u4e2a\u5411\u91cf":1,"\u4ee3\u8868\u4e0d\u540c\u7684\u53cc\u5c42\u5e8f\u5217":1,"\u4ee3\u8868\u5355\u5c42\u5e8f\u5217":1,"\u4ee3\u8868\u53cc\u5c42\u5e8f\u5217":1,"\u4ee4\u884c\u53c2\u6570\u6587\u6863":13,"\u4ee5\u53ca\u53cc\u5c42\u5e8f\u5217":0,"\u4ee5\u53ca\u8ba1\u7b97\u903b\u8f91\u5728\u5e8f\u5217\u4e0a\u7684\u5faa\u73af\u5c55\u5f00":2,"\u4ee5\u592a\u7f51\u5361":9,"\u4ee5\u76f8\u5bf9\u8def\u5f84\u5f15\u7528":22,"\u4ef7\u683c":1,"\u4efb\u610f\u4e00\u79cdcblas\u5b9e\u73b0":4,"\u4f1a\u5bf9\u8fd9\u7c7b\u8f93\u5165\u8fdb\u884c\u62c6\u89e3":2,"\u4f1a\u5c06\u6bcf\u4e2a\u65f6\u95f4\u6b65\u7684\u8f93\u51fa\u62fc\u63a5":2,"\u4f1a\u6210\u4e3astep\u51fd\u6570\u7684\u8f93\u5165":2,"\u4f1a\u62a5\u5bfb\u627e\u4e0d\u5230\u8fd9\u4e9b\u52a8\u6001\u5e93":10,"\u4f1a\u62a5\u9519":2,"\u4f1a\u6839\u636e\u547d\u4ee4\u884c\u53c2\u6570\u6307\u5b9a\u7684\u6d4b\u8bd5\u65b9\u5f0f":22,"\u4f1a\u6839\u636einput_types\u68c0\u67e5\u6570\u636e\u7684\u5408\u6cd5\u6027":23,"\u4f1a\u751f\u6210\u591a\u4e2agener":23,"\u4f1a\u88ab\u62c6\u89e3\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":2,"\u4f1a\u88ab\u62c6\u89e3\u4e3a\u975e\u5e8f\u5217":2,"\u4f1a\u9884\u5148\u8bfb\u53d6\u5168\u90e8\u6570\u636e\u5230\u5185\u5b58\u4e2d":23,"\u4f20\u5165\u547d\u4ee4\u884c\u53c2\u6570\u521d\u59cb\u5316":26,"\u4f20\u5165\u7684\u662f\u4e00\u4e2a\u51fd\u6570":23,"\u4f20\u5165\u7684\u914d\u7f6e\u53c2\u6570\u5305\u62ec":3,"\u4f20\u5165\u8fd9\u4e2a\u53d8\u91cf\u7684\u65b9\u5f0f\u4e3a":23,"\u4f46\u4ece\u4e0a\u9762\u7684\u6570\u636e\u683c\u5f0f\u53ef\u77e5":1,"\u4f46\u5b50\u53e5\u542b\u6709\u7684\u8bcd\u8bed\u6570\u53ef\u4ee5\u4e0d\u76f8\u7b49":2,"\u4f46\u662f":[3,23],"\u4f46\u662f\u5728":23,"\u4f46\u662f\u5982\u679c\u4f7f\u7528\u4e86\u9ad8\u6027\u80fd\u7684\u7f51\u5361":9,"\u4f46\u662f\u65b9\u4fbf\u8c03\u8bd5\u548cbenchmark":4,"\u4f46\u662f\u6709\u65f6\u4e3a\u4e86\u8ba1\u7b97\u5747\u8861\u6027":23,"\u4f46\u7406\u8bba\u4e0a\u652f\u6301\u5176\u4ed6\u7684":10,"\u4f46\u8fd9\u79cd\u60c5\u51b5\u4e0b":1,"\u4f46\u9700\u8981\u6ce8\u610f\u7684\u662f\u7f16\u8bd1\u548c":4,"\u4f4d\u7f6e":1,"\u4f4e\u4e8edocker":3,"\u4f4f":1,"\u4f53\u53ef\u4ee5\u53c2\u8003":23,"\u4f5c\u4e3a\u4e0b\u4e00\u4e2a\u5b50\u53e5memory\u7684\u521d\u59cb\u72b6\u6001":1,"\u4f5c\u4e3a\u5f53\u524d\u65f6\u523b\u8f93\u5165":2,"\u4f5c\u4e3aboot":1,"\u4f5c\u7528":0,"\u4f7f\u5728python\u73af\u5883\u4e0b\u7684\u9884\u6d4b\u63a5\u53e3\u66f4\u52a0\u7b80\u5355":26,"\u4f7f\u7528":[2,4,9,26],"\u4f7f\u7528\u5982\u4e0b\u811a\u672c\u53ef\u4ee5\u786e\u5b9a\u672c\u673a\u7684cpu\u77e5\u5426\u652f\u6301":9,"\u4f7f\u7528\u7684\u547d\u4ee4\u4e5f\u662f":4,"\u4f7f\u7528\u8005\u53ea\u9700\u8981\u5173\u6ce8\u4e8e\u8bbe\u8ba1rnn\u5728\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185\u5b8c\u6210\u7684\u8ba1\u7b97":2,"\u4f7f\u7528\u8be5\u63a5\u53e3\u7528\u6237\u53ef\u4ee5\u53ea\u5173\u6ce8\u5982\u4f55":23,"\u4f7f\u7528\u8be5dockerfile\u6784\u5efa\u51fa\u955c\u50cf":9,"\u4f7f\u7528\u8fd9\u4e2a\u5173\u952e\u8bcd":23,"\u4f7f\u7528deb\u5305\u5728ubuntu\u4e0a\u5b89\u88c5paddlepaddl":8,"\u4f7f\u7528dockerfile\u6784\u5efa\u4e00\u4e2a\u5168\u65b0\u7684dock":9,"\u4f7f\u7528mnist\u624b\u5199\u8bc6\u522b\u4f5c\u4e3a\u6837\u4f8b":26,"\u4f7f\u7528ssh\u8bbf\u95eepaddlepaddle\u955c\u50cf":9,"\u4f86":1,"\u4f8b\u5982":[4,13,23],"\u4f8b\u5982\u6587\u4ef6\u540d\u662f":23,"\u4f8b\u5982rdma\u7f51\u5361":9,"\u4f8b\u5982sigmoid\u53d8\u6362":13,"\u4f9d\u6b21\u8fd4\u56de\u4e86\u6587\u4ef6\u4e2d\u7684\u6bcf\u6761\u6570\u636e":23,"\u4f9d\u7136\u4fdd\u6301\u6bcf\u4e2asubseq\u6700\u540e\u4e00\u4e2a\u5143\u7d20\u7684\u503c\u4e0d\u53d8":1,"\u4fbf\u5229":1,"\u4fbf\u5b9c":1,"\u4fe1\u606f":9,"\u505c\u7535":1,"\u5143\u7d20":0,"\u5143\u7d20\u4e4b\u95f4\u7684\u987a\u5e8f\u662f\u91cd\u8981\u7684\u8f93\u5165\u4fe1\u606f":0,"\u5168\u5bb6":1,"\u5173\u4e8edataprovider\u4e2dinput":1,"\u5173\u4e8eunbound":2,"\u5173\u95edcontain":9,"\u5176\u4e2d":[3,9,10,22,23,26],"\u5176\u4e2d\u6587\u672c\u8f93\u5165\u7c7b\u578b\u5b9a\u4e49\u4e3a\u6574\u6570\u65f6\u5e8f\u7c7b\u578binteg":13,"\u5176\u4e2d\u6bcf\u4e2a\u5143\u7d20\u662f\u53cc\u5c42\u5e8f\u5217\u4e2d\u6bcf\u4e2asubseq\u6700\u540e\u4e00\u4e2a":0,"\u5176\u4e2d\u7b2c\u4e00\u884c\u662f\u5f15\u5165paddlepaddle\u7684pydataprovider2\u5305":23,"\u5176\u4e2d\u7b2ci\u4e2asubseq\u4e2d\u7684\u6240\u6709\u5411\u91cf\u5747\u4e3a\u8f93\u5165\u7684\u5355\u5c42\u5e8f\u5217\u4e2d\u7684\u7b2ci\u4e2a\u5411\u91cf":1,"\u5176\u4ed6\u53c2\u6570\u5747\u4f7f\u7528kei":23,"\u5176\u4ed6\u53c2\u6570\u8bf7\u53c2\u8003":13,"\u5176\u4ed6\u53c2\u6570\u90fd\u4f7f\u7528kei":23,"\u5176\u4f5c\u7528\u662f\u5c06\u8bad\u7ec3\u6570\u636e\u4f20\u5165\u5185\u5b58\u6216\u8005\u663e\u5b58":22,"\u5176\u5b83\u90e8\u5206\u548c\u903b\u8f91\u56de\u5f52\u7f51\u7edc\u7ed3\u6784\u4e00\u81f4":13,"\u5176\u5b83layer\u7684\u8f93\u51fa":2,"\u5176\u6570\u636e\u4f7f\u7528":23,"\u5176\u6b21":1,"\u5176\u7b2c\u4e00\u884c\u8bf4\u660e\u4e86paddle\u7684\u7248\u672c":21,"\u5177":23,"\u5177\u4f53\u53ef\u4ee5\u8bbe\u7f6e\u6210\u4ec0\u4e48\u5176\u4ed6\u683c":23,"\u5177\u4f53\u53ef\u53c2\u8003\u6587\u6863":2,"\u5177\u4f53\u6709\u54ea\u4e9b\u683c\u5f0f":23,"\u5177\u4f53\u8bf7\u53c2\u8003\u6ce8\u610f\u4e8b\u9879\u4e2d\u7684":9,"\u5177\u4f53dataprovider\u8fd8\u5177\u6709\u4ec0\u4e48\u529f\u80fd":23,"\u5177\u6709\u4e24\u4e2a\u53c2\u6570":23,"\u5177\u6709\u548c\u5bbf\u4e3b\u673a\u76f8\u8fd1\u7684\u8fd0\u884c\u6548\u7387":9,"\u5177\u6709\u7684\u5c5e\u6027\u6709":23,"\u5178\u578b\u7684\u8f93\u51fa\u7ed3\u679c\u4e3a":26,"\u5178\u578b\u7684\u9884\u6d4b\u4ee3\u7801\u5982\u4e0b":26,"\u5185\u5b58\u4e0d\u591f\u7528\u7684\u60c5\u51b5":22,"\u5185\u5c42\u662f":1,"\u5185\u5c42inner":1,"\u518d\u6307\u5b9a":4,"\u5199\u5165train":23,"\u5199\u5728train":22,"\u51c6\u5907":1,"\u51c6\u5907\u6570\u636e":26,"\u51fa\u53bb\u73a9":1,"\u51fa\u5dee":1,"\u51fa\u6765":1,"\u51fd\u6570":23,"\u51fd\u6570\u4e2d":23,"\u51fd\u6570\u4e2d\u4f7f\u7528":23,"\u51fd\u6570\u4e2d\u7684":23,"\u51fd\u6570\u53ea\u5173\u6ce8\u4e8ernn\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185\u7684\u8ba1\u7b97":2,"\u51fd\u6570\u5fc5\u987b\u8fd4\u56de\u4e00\u4e2a\u6216\u591a\u4e2alayer\u7684\u8f93\u51fa":2,"\u51fd\u6570\u662f\u4f7f\u7528":23,"\u51fd\u6570\u6765\u4fdd\u8bc1\u517c\u5bb9\u6027":23,"\u51fd\u6570\u67e5\u8be2\u6587\u6863":26,"\u5206\u522b\u4e3a":1,"\u5206\u522b\u4ece\u8bcd\u8bed\u548c\u53e5\u5b50\u7ea7\u522b\u7f16\u7801\u8f93\u5165\u6570\u636e":2,"\u5206\u522b\u5b9a\u4e49\u5b50\u53e5\u7ea7\u522b\u548c\u8bcd\u8bed\u7ea7\u522b\u4e0a\u9700\u8981\u5b8c\u6210\u7684\u8fd0\u7b97":2,"\u5206\u522b\u662f":0,"\u5206\u5e03\u5f0f\u8bad\u7ec3":13,"\u5206\u6790\u5f97\u51fa":1,"\u5206\u7c7b\u6210\u6b63\u9762\u60c5\u7eea\u548c":23,"\u5217\u8868\u5982\u4e0b":23,"\u5219\u53ef\u4ee5\u4f7f\u7528":9,"\u5219\u53ef\u4ee5\u9009\u62e9\u4e0a\u8868\u4e2d\u7684avx\u7248\u672cpaddlepaddl":9,"\u5219\u5b57\u4e0e\u5b57\u4e4b\u95f4\u7528\u7a7a\u683c\u5206\u9694":13,"\u5219\u9700\u8981\u4f7f\u7528":10,"\u5219\u9700\u8981\u5148\u5c06":9,"\u5219\u9700\u8981\u8fdb\u884c\u4e00\u5b9a\u7684\u4e8c\u6b21\u5f00\u53d1":9,"\u521b\u5efa\u4e00\u4e2a":26,"\u521b\u5efagener":23,"\u521d\u59cb\u72b6\u6001":2,"\u5220\u9664contain":9,"\u5229\u7528\u5355\u8bcdid\u67e5\u627e\u5bf9\u5e94\u7684\u8be5\u5355\u8bcd\u7684\u8fde\u7eed\u8868\u793a\u5411\u91cf":13,"\u5229\u7528\u8fd9\u79cd\u7279\u6027":2,"\u5229\u7528\u903b\u8f91\u56de\u5f52\u6a21\u578b\u5bf9\u8be5\u5411\u91cf\u8fdb\u884c\u5206\u7c7b":13,"\u5229\u843d":1,"\u522b":13,"\u5237\u7259":1,"\u524d\u53f0":1,"\u5269\u4e0b\u7684pass\u4f1a\u76f4\u63a5\u4ece\u5185\u5b58\u91cc":23,"\u52a0\u4e86l2\u6b63\u5219\u548c\u68af\u5ea6\u622a\u65ad":13,"\u52a0\u8f7d\u6570\u636e":13,"\u5305":9,"\u5305\u548c":9,"\u5305\u62ec":13,"\u5305\u62ec\u7b80\u5355\u7684rnn\u6a21\u578b":13,"\u5305\u62ecdocker\u955c\u50cf":8,"\u5305\u62ecpaddle\u7684\u4e8c\u8fdb\u5236":9,"\u5305\u62ecpaddle\u8fd0\u884cdemo\u6240\u9700\u8981\u7684\u4f9d\u8d56":9,"\u5341\u4e00":1,"\u534e\u6da6\u4e07\u5bb6":1,"\u5355\u53cc\u5c42\u5e8f\u5217\u7684\u53e5\u5b50\u662f\u4e00\u6837\u7684":1,"\u5355\u53cc\u5c42\u5e8f\u5217\u7684label\u90fd\u5206\u522b\u662f0\u548c1":1,"\u5355\u5c42":2,"\u5355\u5c42\u5e8f\u5217":[0,1],"\u5355\u5c42\u5e8f\u5217\u7684\u6570\u636e":1,"\u5355\u5c42\u5e8f\u5217\u7684\u6bcf\u4e2a\u5143\u7d20":0,"\u5355\u5c42\u5e8f\u5217\u7684dataprovider\u5982\u4e0b":1,"\u5355\u5c42\u5e8f\u5217\u76f4\u63a5\u53d6\u4e86\u6700\u540e\u4e00\u4e2a\u5143\u7d20":1,"\u5355\u5c42\u5e8f\u5217\u7b2ci\u4e2a\u5143\u7d20":0,"\u5355\u5c42\u5e8f\u5217\u8fc7\u4e86\u4e00\u4e2amix":1,"\u5355\u5c42\u6216\u53cc\u5c42":0,"\u5355\u5c42rnn":[1,2],"\u5355\u5c42rnn\u793a\u4f8b":14,"\u5355\u6d4b\u4e2d":1,"\u5355\u8fdb\u5355\u51fa":2,"\u536b\u751f":1,"\u5373":[9,13],"\u5373\u4e00\u4e2a\u53e5\u5b50\u4e00\u4e2alabel":1,"\u5373\u4e00\u4e2a\u5b50\u53e5\u4e00\u4e2alabel":1,"\u5373\u4e0d\u5728\u4e4e\u5185\u5b58\u6682\u5b58\u591a\u5c11\u6761\u6570\u636e":23,"\u5373\u4e0d\u662f\u4e00\u6761\u5e8f\u5217":23,"\u5373\u4ece\u5355\u8bcd\u5b57\u7b26\u4e32\u5230\u5355\u8bcdid\u7684\u5b57\u5178":23,"\u5373\u4f1a\u751f\u6210100\u4e2agener":23,"\u5373\u4f7f\u5728check\u4e2d\u6570\u636e\u4e0d\u5408\u6cd5":23,"\u5373\u4f7f\u5728process\u91cc\u9762\u53ea\u4f1a\u6709\u4e00":23,"\u5373\u5185\u5c42memory\u7684boot":1,"\u5373\u521d\u59cb\u72b6\u6001\u4e3a0":2,"\u5373\u5305\u542b\u65f6\u95f4\u6b65\u4fe1\u606f":23,"\u5373\u53cc\u5c42rnn\u7684\u6bcf\u4e2a\u72b6\u6001":2,"\u5373\u53ef":23,"\u5373\u53ef\u4ee5\u4f7f\u7528ssh\u8bbf\u95ee\u5bbf\u4e3b\u673a\u76848022\u7aef\u53e3":9,"\u5373\u53ef\u542f\u52a8\u548c\u8fdb\u5165paddlepaddle\u7684contain":9,"\u5373\u53ef\u5728\u672c\u5730\u7f16\u8bd1\u51fapaddlepaddle\u7684\u955c\u50cf":3,"\u5373\u53ef\u6253\u5370\u51fapaddlepaddle\u7684\u7248\u672c\u548c\u6784\u5efa":9,"\u5373\u5927\u90e8\u5206\u503c\u4e3a0":23,"\u5373\u5982\u679ctrain":23,"\u5373\u5bf9\u7b2c3\u6b65\u8fdb\u884c\u66ff\u6362":13,"\u5373\u628a\u5355\u5c42rnn\u751f\u6210\u540e\u7684subseq\u7ed9\u62fc\u63a5\u6210\u4e00\u4e2a\u65b0\u7684\u53cc\u5c42seq":2,"\u5373\u6574\u4e2a\u53cc\u5c42group\u662f\u5c06\u524d\u4e00\u4e2a\u5b50\u53e5\u7684\u6700\u540e\u4e00\u4e2a\u5411\u91cf":1,"\u5373\u6574\u4e2a\u8f93\u5165\u5e8f\u5217":0,"\u5373\u662f\u4e00\u6761\u65f6\u95f4\u5e8f\u5217":23,"\u5373\u8d77\u5230\u7684\u4f5c\u7528\u4ec5\u4ec5\u662f\u628a\u53cc\u5c42seq\u62c6\u6210\u5355\u5c42":1,"\u5373input":2,"\u5373train":23,"\u5377\u79ef\u7f51\u7edc\u662f\u4e00\u79cd\u7279\u6b8a\u7684\u4ece\u8bcd\u5411\u91cf\u8868\u793a\u5230\u53e5\u5b50\u8868\u793a\u7684\u65b9\u6cd5":13,"\u53bb\u8fc7":1,"\u53c2\u6570":3,"\u53c2\u6570\u6570\u91cf":13,"\u53c2\u8003":22,"\u53c2\u89c1":[6,7],"\u53c2\u89c1pydataprovider2":1,"\u53c8":1,"\u53c8\u662f\u4e00\u4e2a\u5355\u5c42\u7684\u5e8f\u5217":0,"\u53cc\u5c42":2,"\u53cc\u5c42\u5e8f\u5217":[0,1],"\u53cc\u5c42\u5e8f\u5217\u5728\u540c\u6837\u7684mix":1,"\u53cc\u5c42\u5e8f\u5217\u662f\u4e00\u4e2a\u5d4c\u5957\u7684\u5e8f\u5217":0,"\u53cc\u5c42\u5e8f\u5217\u662fpaddlepaddle\u652f\u6301\u7684\u4e00\u79cd\u975e\u5e38\u7075\u6d3b\u7684\u6570\u636e\u7ec4\u7ec7\u65b9\u5f0f":2,"\u53cc\u5c42\u5e8f\u5217\u6bcf\u4e2asubseq\u4e2d\u6bcf\u4e2a\u5143\u7d20":0,"\u53cc\u5c42\u5e8f\u5217\u7684\u6570\u636e":1,"\u53cc\u5c42\u5e8f\u5217\u7684dataprovider\u5982\u4e0b":1,"\u53cc\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":0,"\u53cc\u5c42\u5e8f\u5217\u9996\u5148":1,"\u53cc\u5c42\u6216\u8005\u5355\u5c42":0,"\u53cc\u5c42rnn":[1,2],"\u53cc\u5c42rnn\u793a\u4f8b":14,"\u53cc\u8fdb\u5355\u51fa":2,"\u53d1\u884c\u7248":10,"\u53d6\u4e86\u6bcf\u4e2asubseq\u7684\u5e73\u5747\u503c":1,"\u53d6\u4e86\u6bcf\u4e2asubseq\u7684\u6700\u540e\u4e00\u4e2a\u5143\u7d20":1,"\u53d6\u51b3\u4e8e\u662f\u5426\u5bfb\u627e\u5230gflags":4,"\u53d6\u51b3\u4e8e\u662f\u5426\u5bfb\u627e\u5230glog":4,"\u53d6\u51b3\u4e8e\u662f\u5426\u5bfb\u627e\u5230gtest":4,"\u53d6\u51b3\u4e8e\u662f\u5426\u627e\u5230swig":4,"\u53d8\u4e3a3\u4e2a\u65b0\u7684\u5b50\u6b65\u9aa4":13,"\u53d8\u4f1a\u62a5\u8fd9\u4e2a\u9519\u8bef":10,"\u53d8\u91cf":23,"\u53e3\u5934":1,"\u53e5\u5b50\u8868\u793a\u7684\u8ba1\u7b97\u66f4\u65b0\u4e3a2\u6b65":13,"\u53ea\u4f5c\u4e3aread":2,"\u53ea\u5305\u62ecpaddle\u7684\u4e8c\u8fdb\u5236":9,"\u53ea\u662f\u53cc\u5c42\u5e8f\u5217\u5c06\u5176\u53c8\u505a\u4e86\u5b50\u5e8f\u5217\u5212\u5206":1,"\u53ea\u662f\u5c06\u53e5\u5b50\u5229\u7528\u8fde\u7eed\u5411\u91cf\u8868\u793a\u66ff\u6362\u7a00\u758f":13,"\u53ea\u662f\u8bf4\u660e\u6570\u636e\u7684\u987a\u5e8f\u662f\u91cd\u8981\u7684":23,"\u53ea\u6709":1,"\u53ea\u7528\u4e8e\u5728\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u4e2d\u6307\u5b9a\u8f93\u5165\u6570\u636e":2,"\u53ea\u80fd\u591f\u8fd4\u56delist\u6216\u8005tupl":23,"\u53ea\u80fd\u901a\u8fc7":1,"\u53ea\u8bfbmemory\u8f93\u5165":2,"\u53ea\u9700\u8981\u4f7f\u7528\u4e00\u884c\u4ee3\u7801\u5373\u53ef\u4ee5\u8bbe\u7f6e\u8bad\u7ec3\u5f15\u7528\u8fd9\u4e2adataprovid":23,"\u53ea\u9700\u8981\u5728":23,"\u53ea\u9700\u8981\u77e5\u9053\u8fd9\u53ea\u662f\u4e00\u4e2a\u6807\u8bb0\u5c5e\u6027\u7684\u65b9\u6cd5\u5c31\u53ef\u4ee5\u4e86":23,"\u53ef\u4ee5":1,"\u53ef\u4ee5\u4e3a\u4e00\u4e2a\u6570\u636e\u6587\u4ef6\u8fd4\u56de\u591a\u6761\u8bad\u7ec3\u6837\u672c":23,"\u53ef\u4ee5\u4f20\u516510k":3,"\u53ef\u4ee5\u4f7f\u7528":3,"\u53ef\u4ee5\u4f7f\u7528\u547d\u4ee4":10,"\u53ef\u4ee5\u4f7f\u7528\u8bad\u7ec3\u597d\u7684\u6a21\u578b\u8bc4\u4f30\u5e26\u6709label\u7684\u9a8c\u8bc1\u96c6":13,"\u53ef\u4ee5\u4f7f\u7528graphviz\u5bf9paddlepaddle\u7684\u7f51\u7edc\u6a21\u578b\u8fdb\u884c\u7ed8\u5236":16,"\u53ef\u4ee5\u4f7f\u7528paddl":16,"\u53ef\u4ee5\u4f7f\u7528python\u7684":26,"\u53ef\u4ee5\u53c2\u8003":13,"\u53ef\u4ee5\u5728\u4e00\u4e2a\u51fd\u6570\u91cc":23,"\u53ef\u4ee5\u5728cmake\u7684\u547d\u4ee4\u884c\u8bbe\u7f6e":4,"\u53ef\u4ee5\u5c06\u4e00\u6761\u6570\u636e\u8bbe\u7f6e\u6210\u591a\u4e2abatch":23,"\u53ef\u4ee5\u5c06memory\u7406\u89e3\u4e3a\u4e00\u4e2a\u65f6\u5ef6\u64cd\u4f5c":2,"\u53ef\u4ee5\u5c06paddlepaddle\u7684\u6a21\u578b\u548c\u914d\u7f6e\u6253\u5305\u6210\u4e00\u4e2a\u6587\u4ef6":16,"\u53ef\u4ee5\u5c06paddlepaddle\u7684\u8bad\u7ec3\u6a21\u578b\u4ee5proto":16,"\u53ef\u4ee5\u65b9\u4fbf\u5d4c\u5165\u5f0f\u5de5\u4f5c":4,"\u53ef\u4ee5\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":2,"\u53ef\u4ee5\u662f\u4e00\u4e2a\u975e\u5e8f\u5217":2,"\u53ef\u4ee5\u663e\u793a\u5730\u6307\u5b9a\u4e00\u4e2alayer\u7684\u8f93\u51fa\u7528\u4e8e\u521d\u59cb\u5316memori":2,"\u53ef\u4ee5\u6709\u4ee5\u4e0b\u4e24\u79cd":2,"\u53ef\u4ee5\u6839\u636e\u4e0d\u540c\u7684\u6570\u636e\u914d\u7f6e\u4e0d\u540c\u7684\u8f93\u5165\u7c7b\u578b":23,"\u53ef\u4ee5\u770b\u4f5c\u662f\u4e00\u4e2a\u975e\u5e8f\u5217\u8f93\u5165":0,"\u53ef\u4ee5\u8fd4\u56de\u4e00\u4e2adict":23,"\u53ef\u4ee5\u901a\u8fc7show":13,"\u53ef\u7528\u5728\u6d4b\u8bd5\u6216\u8bad\u7ec3\u65f6\u6307\u5b9a\u521d\u59cb\u5316\u6a21\u578b":13,"\u53ef\u80fd\u7684\u5185\u5b58\u6cc4\u9732\u95ee\u9898":22,"\u53ef\u80fd\u7684\u8f93\u51fa\u4e3a":10,"\u53ef\u9009":23,"\u5403":1,"\u5403\u996d":1,"\u5404\u65b9\u9762":1,"\u5404\u79cd\u53c2\u6570\u548c\u7ef4\u62a4":3,"\u5408":1,"\u5408\u7406":1,"\u540c\u65f6":[3,4,23],"\u540c\u65f6\u4e5f\u80fd\u591f\u5f15\u5165\u66f4\u52a0\u590d\u6742\u7684\u8bb0\u5fc6\u673a\u5236":2,"\u540c\u65f6\u4f1a\u8ba1\u7b97\u5206\u7c7b\u51c6\u786e\u7387":13,"\u540c\u65f6\u6b22\u8fce\u8d21\u732e\u66f4\u591a\u7684\u5b89\u88c5\u5305":8,"\u540c\u6837\u53ef\u4ee5\u6269\u5c55\u5230\u53cc\u5c42\u5e8f\u5217\u7684\u5904\u7406\u4e0a":2,"\u540d\u79f0":13,"\u540e\u9762\u8ddf\u7740\u4e00\u7cfb\u5217\u7f16\u8bd1\u53c2\u6570":21,"\u5411\u91cf\u8868\u793a":13,"\u5426":4,"\u5426\u5219":22,"\u5426\u5219\u5728\u7b2c0\u4e2a\u65f6\u95f4\u6b65\u65f6":1,"\u5426\u5219\u9700\u8981\u9009\u62e9\u975eavx\u7684paddlepaddl":9,"\u5440":1,"\u5468\u56f4":1,"\u547d\u4ee4":3,"\u547d\u4ee4\u4e3a":9,"\u547d\u4ee4\u5373\u53ef\u5b8c\u6210\u5b89\u88c5":10,"\u547d\u4ee4\u6307\u5b9a\u7684\u53c2\u6570\u4f1a\u4f20\u5165\u7f51\u7edc\u914d\u7f6e\u4e2d":13,"\u547d\u4ee4\u8fd0\u884c\u955c\u50cf":9,"\u547d\u4ee4\u9884\u5148\u4e0b\u8f7d\u955c\u50cf":9,"\u548c\u4e00\u4e2a\u5df2\u7ecf\u5206\u8bcd\u540e\u7684\u53e5\u5b50":1,"\u548c\u4e09\u79cd\u5e8f\u5217\u6a21\u5f0f":23,"\u548c\u53cc\u5c42\u5e8f\u5217\u542b\u6709subseq":0,"\u548c\u53cc\u5c42rnn":1,"\u548c\u5dee\u8bc4":13,"\u548c\u5e8f\u5217\u4e2d\u542b\u6709\u5143\u7d20\u7684\u6570\u76ee\u540c":0,"\u548c\u6587\u672c\u4fe1\u606f\u7528tab\u95f4\u9694":13,"\u548c\u6d4b\u8bd5\u6587\u4ef6\u5217\u8868":22,"\u548c\u7528\u6237\u4f20\u5165\u7684\u53c2\u6570":23,"\u548c\u90e8\u5206layer":2,"\u548c\u9884\u5904\u7406\u811a\u672c":13,"\u548cavgpool":0,"\u548ccudnn":10,"\u548cinitalizer\u91cc\u5b9a\u4e49\u987a\u5e8f\u4e00\u81f4":13,"\u54c1\u8d28":1,"\u5546\u52a1":1,"\u554a":1,"\u5668":13,"\u56db\u4e2a\u7248\u672c":10,"\u56db\u79cd\u6570\u636e\u7c7b\u578b\u662f":23,"\u56e0\u6b64":2,"\u56e0\u6b642\u4e2abatch\u5c31\u53ef\u4ee5\u5b8c\u62101\u4e2apass":1,"\u56e0\u6b64\u4e0a\u8ff0\u4e09\u4e2alayer\u7684\u524d\u5411\u4f1a\u62a5\u51fa":1,"\u56e0\u6b64\u4e24\u4e2a\u914d\u7f6e\u5728\u8fd9\u4e24\u5c42\u4e0a\u7684\u8f93\u51fa\u662f\u4e00\u6837\u7684":1,"\u56e0\u6b64\u5355\u53cc\u5c42\u5e8f\u5217\u7684\u8f93\u51fa\u662f\u4e00\u6837\u65f3":1,"\u56e0\u6b64\u53cc\u5c42\u5e8f\u5217\u7684\u914d\u7f6e\u4e2d":1,"\u56e0\u6b64\u53cc\u5c42\u5e8f\u5217\u8fc7\u5b8clstmemory\u7684\u8f93\u51fa\u548c\u5355\u5c42\u7684\u4e00\u6837":1,"\u56e0\u6b64\u5f53\u5916\u5c42\u6709i":1,"\u56fe\u50cf\u5206\u7c7b":12,"\u5728":[0,1,4,10,23],"\u5728\u58f0\u660edataprovider\u7684\u65f6\u5019\u4f20\u5165\u4e86dictionary\u4f5c\u4e3a\u53c2\u6570":23,"\u5728\u5b8c\u6210\u4e86\u6570\u636e\u548c\u7f51\u7edc\u7ed3\u6784\u642d\u5efa\u4e4b\u540e":13,"\u5728\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u4e2d":2,"\u5728\u672c\u95ee\u9898\u4e2d":13,"\u5728\u6a21\u578b\u914d\u7f6e\u4e2d\u5229\u7528":13,"\u5728\u6b64\u4e3a\u65b9\u4fbf\u5bf9\u6bd4\u4e0d\u540c\u7f51\u7edc\u7ed3\u6784":13,"\u5728\u6bcf\u4e2a\u7ef4\u5ea6\u4e0a\u53d6\u51fa\u5728\u8be5\u53e5\u8bdd\u65b0\u7684\u5411\u91cf\u96c6\u5408\u4e0a\u8be5\u7ef4\u5ea6\u7684\u6700\u5927\u503c\u4f5c\u4e3a\u6700\u540e\u7684\u53e5\u5b50\u8868\u793a\u5411\u91cf":13,"\u5728\u7a0b\u5e8f\u5f00\u59cb\u9636\u6bb5":26,"\u5728\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4efb\u52a1\u4e2d":0,"\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u8fdb\u884c\u6d4b\u8bd5":22,"\u5728\u8bad\u7ec3\u914d\u7f6e\u91cc":23,"\u5728\u8f93\u51fa\u7684\u8fc7\u7a0b\u4e2d":2,"\u5728\u8fd9\u4e2a\u51fd\u6570\u4e2d":23,"\u5728\u8fd9\u79cd\u7ed3\u6784\u4e2d":2,"\u5728\u8fd9\u91cc":2,"\u5728\u914d\u7f6e\u4e2d\u8bfb\u53d6\u4e86\u5b57\u5178":23,"\u5728cmake\u914d\u7f6e\u65f6\u53ef\u4ee5\u4f7f\u7528":4,"\u5728paddlepaddle\u4e2d":2,"\u5728pydataprovider\u4e2d":23,"\u5728python\u73af\u5883\u4e0b\u9884\u6d4b\u7ed3\u679c":26,"\u5728step\u51fd\u6570\u4e2d\u5b9a\u4e49":2,"\u5728step\u51fd\u6570\u4e2d\u5b9a\u4e49memori":2,"\u5730\u6bb5":1,"\u5730\u7406\u4f4d\u7f6e":1,"\u5730\u94c1\u7ad9":1,"\u57fa\u4e8e\u53cc\u5c42\u5e8f\u5217\u8f93\u5165":2,"\u57fa\u672c\u4e0a\u4e0d\u80fd\u6574\u4f53\u4fee\u6b63":23,"\u57fa\u672c\u7684\u5904\u7406\u903b\u8f91\u4e5f\u548cmnist\u903b\u8f91\u4e00\u81f4":23,"\u57fa\u672c\u7684pydataprovider\u4f7f\u7528\u4ecb\u7ecd\u5b8c\u6bd5\u4e86":23,"\u5904\u7406\u7684\u8f93\u5165\u5e8f\u5217\u4e3b\u8981\u5206\u4e3a\u4ee5\u4e0b\u4e09\u79cd\u7c7b\u578b":2,"\u5916\u5c42memory\u5fc5\u987b\u6709boot":1,"\u5916\u5c42memory\u662f\u4e00\u4e2a\u5143\u7d20":1,"\u5916\u5c42memory\u662f\u5355\u5c42\u5e8f\u5217":1,"\u5916\u5c42outer":1,"\u591a\u4e2ainput\u4ee5list\u65b9\u5f0f\u8f93\u5165":13,"\u591a\u53e5\u8bdd\u8fdb\u4e00\u6b65\u6784\u6210\u4e86\u6bb5\u843d":2,"\u591a\u6b21\u8fd4\u56de\u53d8\u91cf":23,"\u591a\u7ebf\u7a0b\u4e0b\u8f7d\u8fc7\u7a0b\u4e2d":3,"\u591a\u7ebf\u7a0b\u6570\u636e\u8bfb\u53d6":23,"\u591a\u8f6e\u5bf9\u8bdd\u7b49\u66f4\u4e3a\u590d\u6742\u7684\u8bed\u8a00\u6570\u636e":2,"\u5927":23,"\u5929":1,"\u5929\u4e00\u5e7f\u573a":1,"\u5929\u4e00\u9601":1,"\u597d":1,"\u597d\u5403":1,"\u597d\u8bc4":13,"\u5982\u4e0b":1,"\u5982\u679c":[10,23],"\u5982\u679c\u4e0d\u4e86\u89e3":23,"\u5982\u679c\u4e0d\u4f7f\u7528\u5219\u4f1a\u4f7f\u7528\u4e00\u4e2a\u7b80\u5316\u7248\u7684\u547d\u4ee4\u884c\u53c2\u6570\u89e3\u6790":4,"\u5982\u679c\u4e0d\u4f7f\u7528\u5219\u4f1a\u4f7f\u7528\u4e00\u4e2a\u7b80\u5316\u7248\u7684\u65e5\u5fd7\u5b9e\u73b0":4,"\u5982\u679c\u4e0d\u5207\u8bcd":13,"\u5982\u679c\u4e0d\u8bbe\u7f6e\u7684\u8bdd":23,"\u5982\u679c\u4f7f\u7528gpu\u7248\u672c\u7684paddlepaddl":10,"\u5982\u679c\u5185\u5c42memory\u60f3":1,"\u5982\u679c\u5728":10,"\u5982\u679c\u5728\u7b2c\u4e00\u6b21cmake\u4e4b\u540e\u60f3\u8981\u91cd\u65b0\u8bbe":4,"\u5982\u679c\u5728\u8bad\u7ec3\u65f6":23,"\u5982\u679c\u5c0f\u4e8e\u8fd9\u4e2a\u4e0b\u8f7d\u901f\u5ea6":3,"\u5982\u679c\u60a8\u4f7f\u7528":9,"\u5982\u679c\u60f3\u8981\u5728\u5916\u90e8\u673a\u5668\u8bbf\u95ee\u8fd9\u4e2acontain":9,"\u5982\u679c\u662ffalse\u7684\u8bdd":23,"\u5982\u679c\u662ftrue\u7684\u8bdd":23,"\u5982\u679c\u6709\u591a\u4e2a\u8f93\u5165":2,"\u5982\u679c\u6709\u591a\u4e2a\u8f93\u5165\u5e8f\u5217":2,"\u5982\u679c\u6709\u66f4\u590d\u6742\u7684\u4f7f\u7528":22,"\u5982\u679c\u6ca1\u6709\u5b9a\u4e49memori":2,"\u5982\u679c\u7528\u6237\u4e0d\u6307\u5b9a\u8fd4\u56de\u6570\u636e\u7684\u5bf9\u5e94\u5173\u7cfb":23,"\u5982\u679c\u8bbe\u7f6e\u6210true\u7684\u8bdd":23,"\u5982\u679c\u8f93\u51fa":9,"\u5982\u679c\u8fd0\u884cgpu\u7248\u672c\u7684paddlepaddl":9,"\u5982\u679ctest":22,"\u5b50":1,"\u5b50\u53e5":2,"\u5b50\u53e5\u7684\u5355\u8bcd\u6570\u548c\u6307\u5b9a\u7684\u4e00\u4e2a\u8f93\u5165\u5e8f\u5217\u4e00\u81f4":2,"\u5b81\u6ce2":1,"\u5b83\u5305\u542b\u7684\u53c2\u6570\u6709":23,"\u5b83\u7684":1,"\u5b83\u7684\u6bcf\u4e00\u4e2a\u5143\u7d20":0,"\u5b89\u6392":1,"\u5b89\u88c5\u5305\u5728ubuntu":10,"\u5b89\u88c5\u5305\u7684\u4e0b\u8f7d\u5730\u5740\u662f":10,"\u5b89\u88c5\u597d\u7684paddlepaddle\u811a\u672c\u5305\u62ec\u591a\u6761\u547d\u4ee4":16,"\u5b89\u88c5\u5b8c\u6210\u540e":10,"\u5b89\u88c5\u5b8c\u6210\u7684paddlepaddle\u4e3b\u4f53\u5305\u62ec\u4e09\u4e2a\u90e8\u5206":9,"\u5b89\u88c5\u5b8c\u6210paddlepaddle\u540e":10,"\u5b89\u88c5\u6559\u7a0b":13,"\u5b89\u88c5\u65b9\u6cd5\u8bf7\u53c2\u8003":9,"\u5b89\u88c5\u7f16\u8bd1\u4f9d\u8d56":6,"\u5b89\u88c5\u7f16\u8bd1paddlepaddle\u9700\u8981\u7684\u4f9d\u8d56":5,"\u5b89\u88c5docker\u9700\u8981\u60a8\u7684\u673a\u5668":9,"\u5b89\u88c5paddlepaddl":13,"\u5b89\u88c5paddlepaddle\u7684docker\u955c\u50cf":8,"\u5b89\u9759":1,"\u5b8c\u6210\u4efb\u610f\u7684\u8fd0\u7b97\u903b\u8f91":2,"\u5b8c\u6210\u591a\u673a\u8bad\u7ec3":16,"\u5b8c\u6210\u76f8\u5e94\u7684\u8ba1\u7b97":0,"\u5b8c\u6574\u4ee3\u7801\u89c1":26,"\u5b9a\u4e49\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185rnn\u5355\u5143\u5b8c\u6210\u7684\u8ba1\u7b97":2,"\u5b9a\u4e49\u4e86\u4e00\u4e2a\u53ea\u8bfb\u7684memori":2,"\u5b9a\u4e49\u5728\u5916\u5c42":2,"\u5b9a\u4e49\u6587\u672c\u4fe1\u606f":13,"\u5b9e\u73b0\u4e86\u6253\u5f00\u6587\u672c\u6587\u4ef6":23,"\u5b9e\u73b0\u8bcd\u8bed\u548c\u53e5\u5b50\u4e24\u4e2a\u7ea7\u522b\u7684\u53cc\u5c42rnn\u7ed3\u6784":2,"\u5b9e\u9645\u4e2d\u5e76\u4e0d\u9700\u8981":1,"\u5ba2\u6237":1,"\u5bb6":1,"\u5bc6\u7801\u4e5f\u662froot":9,"\u5bf9":1,"\u5bf9\u4e8e\u7528\u6237\u6765\u8bf4":23,"\u5bf9\u4e8e\u7ed9\u5b9a\u7684\u4e00\u6761\u6587\u672c":13,"\u5bf9\u4e8ecuda\u7684toolkit\u6709\u65ad\u70b9\u7eed\u4f20\u548c\u4f20\u8f93\u901f\u5ea6\u8fc7\u5c0f\u91cd\u542f\u4e0b\u8f7d\u7684":3,"\u5bf9\u4e8emnist\u800c\u8a00":23,"\u5bf9\u5e94\u4e00\u4e2a\u5b50\u53e5":2,"\u5bf9\u5e94\u4e00\u4e2a\u8bcd":2,"\u5bf9\u8be5\u8868\u793a\u8fdb\u884c\u975e\u7ebf\u6027\u53d8\u6362":13,"\u5bf9\u8c61":23,"\u5bf9\u8c61convert":26,"\u5bf9\u8f93\u51fa\u7684\u5408\u5e76":2,"\u5bf9\u9762":1,"\u5c06\u4f1a\u6d4b\u8bd5\u914d\u7f6e\u6587\u4ef6\u4e2dtest":13,"\u5c06\u5176\u6269\u5c55\u6210\u4e00\u4e2a\u65b0\u7684\u53cc\u5c42\u5e8f\u5217":1,"\u5c06\u5176\u62fc\u63a5\u6210\u4e00\u4e2a\u65b0\u7684\u5355\u5c42\u5e8f\u5217":1,"\u5c06\u542b\u6709\u5b50\u53e5":2,"\u5c06\u542b\u6709\u8bcd\u8bed\u7684\u53e5\u5b50\u5b9a\u4e49\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":2,"\u5c06\u5b57\u5178\u5b58\u5165\u4e86set":23,"\u5c06\u5bbf\u4e3b\u673a\u76848022\u7aef\u53e3\u6620\u5c04\u5230container\u768422\u7aef\u53e3\u4e0a":9,"\u5c06\u6570\u636e\u7ec4\u5408\u6210batch\u8bad\u7ec3":23,"\u5c06\u6587\u4ef6\u7684\u7edd\u5bf9\u8def\u5f84\u6216\u76f8\u5bf9\u8def\u5f84":22,"\u5c06\u8bc4\u8bba\u5206\u4e3a\u597d\u8bc4":13,"\u5c06\u8be5\u53e5\u8bdd\u5305\u542b\u7684\u6240\u6709\u5355\u8bcd\u5411\u91cf\u6c42\u5e73\u5747\u5f97\u5230\u53e5\u5b50\u7684\u8868\u793a":13,"\u5c06ssh\u88c5\u5165\u7cfb\u7edf\u5185\u5e76\u5f00\u542f\u8fdc\u7a0b\u8bbf\u95ee":9,"\u5c1a\u53ef":1,"\u5c31":[1,23],"\u5c31\u50cf\u8fd9\u4e2a\u6837\u4f8b\u4e00\u6837":23,"\u5c31\u662f":1,"\u5c31\u662f\u5c06\u8fd9\u4e9b\u52a8\u6001\u5e93\u52a0\u5230\u73af\u5883\u53d8\u91cf\u91cc\u9762":10,"\u5c42\u6b21\u5316\u7684rnn":2,"\u5c45\u7136":1,"\u5c5e\u6027":23,"\u5dee\u8bc4":13,"\u5e2e\u52a9\u6211\u4eec\u5b8c\u6210\u5bf9\u8f93\u5165\u5e8f\u5217\u7684\u62c6\u5206":2,"\u5e2e\u52a9\u6211\u4eec\u66f4\u597d\u5730\u63cf\u8ff0\u6bb5\u843d":2,"\u5e2e\u52a9\u6211\u4eec\u6784\u9020\u4e00\u4e9b\u590d\u6742\u7684\u8f93\u5165\u4fe1\u606f":0,"\u5e38\u89c1\u7684\u8f93\u51fa\u683c\u5f0f\u4e3a":21,"\u5e72\u51c0":1,"\u5e76\u4e14":23,"\u5e76\u4e14\u4f7f\u7528\u5173\u952e\u8bcd":23,"\u5e76\u4e14\u5220\u9664container\u4e2d\u7684\u6570\u636e":9,"\u5e76\u4e14\u5728\u5185\u5b58\u8db3\u591f":23,"\u5e76\u4e14\u6807\u8bb0process\u51fd\u6570\u662f\u4e00\u4e2adataprovid":23,"\u5e76\u4f7f\u7528\u4e86dropout":13,"\u5e76\u572823\u884c\u8fd4\u56de\u7ed9paddlepaddle\u8fdb\u7a0b":23,"\u5e76\u5bf9\u5176\u8be6\u7ec6\u5206\u6790":1,"\u5e76\u5c06\u6bcf\u884c\u8f6c\u6362\u6210\u548c":23,"\u5e76\u63d0\u4f9b":9,"\u5e76\u63d0\u4f9b\u4e86\u7b80\u5355\u7684cache\u529f\u80fd":23,"\u5e76\u8bbe\u7f6e\u597d\u5bf9\u5e94\u7684\u73af\u5883\u53d8\u91cf":10,"\u5e76\u9010\u6e10\u5c55\u793a\u66f4\u52a0\u6df1\u5165\u7684\u529f\u80fd":13,"\u5e8a\u4e0a\u7528\u54c1":1,"\u5e8a\u57ab":1,"\u5e8f\u5217\u4e2d\u542b\u6709\u5143\u7d20\u7684\u6570\u76ee\u540clayer2\u4e00\u81f4":0,"\u5e8f\u5217\u6570\u636e\u548c\u4e0a\u9762\u7684\u5b8c\u5168\u4e00\u6837":1,"\u5e8f\u5217\u6570\u636e\u662f\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4efb\u52a1\u9762\u5bf9\u7684\u4e00\u79cd\u4e3b\u8981\u8f93\u5165\u6570\u636e\u7c7b\u578b":2,"\u5e8f\u5217\u662f\u4e00\u79cd\u5e38\u89c1\u7684\u6570\u636e\u7c7b\u578b":0,"\u5e8f\u5217\u6a21\u578b\u6570\u636e\u63d0\u4f9b":22,"\u5e8f\u5217\u6a21\u578b\u662f\u6307\u6570\u636e\u7684\u67d0\u4e00\u7ef4\u5ea6\u662f\u4e00\u4e2a\u5e8f\u5217\u5f62\u5f0f":23,"\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u5927\u591a\u9075\u5faaencod":2,"\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u7684\u8f93\u5165":2,"\u5e8f\u5217\u7684\u6bcf\u4e2a\u5143\u7d20\u662f\u539f\u6765\u53cc\u5c42\u5e8f\u5217\u6bcf\u4e2asubseq\u5143\u7d20\u7684\u5e73\u5747\u503c":0,"\u5e93\u7684\u8bdd":10,"\u5e94\u8be5":1,"\u5f0f":23,"\u5f15\u7528\u7684dataprovider\u662f":23,"\u5f15\u7528memory\u5f97\u5230\u8fd9layer\u4e0a\u4e00\u65f6\u523b\u8f93\u51fa":2,"\u5f3a\u70c8\u63a8\u8350":1,"\u5f53\u51fd\u6570\u8fd4\u56de\u7684\u65f6\u5019":23,"\u5f53\u524d\u7684\u8f93\u5165y\u548c\u4e0a\u4e00\u4e2a\u65f6\u95f4\u6b65\u7684\u8f93\u51farnn":1,"\u5f53\u524dlog_period\u4e2abatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747\u5206\u7c7b\u9519\u8bef\u7387":13,"\u5f53\u524dlog_period\u4e2abatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747cost":13,"\u5f53\u7136":22,"\u5f53\u8c03":23,"\u5f62\u6210recurr":2,"\u5f62\u6210recurrent\u8fde\u63a5":2,"\u5f88":[1,13],"\u5f88\u591a":1,"\u5f88\u5b89\u9759":1,"\u5f88\u5e72\u51c0":1,"\u5f88\u65b9\u4fbf":1,"\u5f97":1,"\u5f97\u5230\u7ed3\u679c":10,"\u5faa\u73af\u5c55\u5f00\u7684\u6bcf\u4e2a\u65f6\u95f4\u6b65\u603b\u662f\u80fd\u591f\u5f15\u7528\u6240\u6709\u8f93\u5165":2,"\u5fc5\u987b\u5c06\u524d\u4e00\u4e2a\u5b50\u53e5\u7684\u6700\u540e\u4e00\u4e2a\u5143\u7d20":1,"\u5fc5\u987b\u6307\u5411\u4e00\u4e2apaddlepaddle\u5b9a\u4e49\u7684lay":2,"\u5fc5\u987b\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":2,"\u5fc5\u987b\u662f\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":0,"\u5feb":1,"\u5feb\u901f\u5165\u95e8":14,"\u5ff5\u662f":23,"\u6027\u4ef7\u6bd4":1,"\u603b\u4f53\u6765\u8bf4":1,"\u60a8\u4e5f\u53ef\u4ee5\u91c7\u7528\u522b\u7684\u7ec4\u7ec7\u5f62\u5f0f":1,"\u60a8\u53ef\u4ee5\u4f7f\u7528":9,"\u60a8\u5c31\u53ef\u4ee5\u8fdc\u7a0b\u7684\u4f7f\u7528paddlepaddle\u5566":9,"\u60a8\u9700\u8981\u5728\u673a\u5668\u4e2d\u5b89\u88c5\u597ddocker":9,"\u60a8\u9700\u8981\u8fdb\u5165\u955c\u50cf\u8fd0\u884cpaddlepaddl":9,"\u60c5\u611f\u5206\u6790":12,"\u60f3\u8981\u8fd0\u884cpaddlepaddl":9,"\u611f\u89c9":1,"\u6210\u4e3a\u7ef4\u5ea6\u4e3ahidden":13,"\u6211\u4eec\u4ece\u63d0\u524d\u7ed9\u5b9a\u7684\u7c7b\u522b\u96c6\u5408\u4e2d\u9009\u62e9\u5176\u6240\u5c5e\u7c7b":13,"\u6211\u4eec\u4ee5\u6587\u672c\u5206\u7c7b\u95ee\u9898\u4f5c\u4e3a\u80cc\u666f":13,"\u6211\u4eec\u4f7f\u7528":13,"\u6211\u4eec\u53ef\u4ee5\u6309\u7167\u5982\u4e0b\u5c42\u6b21\u5b9a\u4e49\u975e\u5e8f\u5217":0,"\u6211\u4eec\u53ef\u4ee5\u8bbe\u8ba1\u642d\u5efa\u4e00\u4e2a\u7075\u6d3b\u7684":2,"\u6211\u4eec\u5728":1,"\u6211\u4eec\u5728\u6b64\u603b":13,"\u6211\u4eec\u5c06\u4ee5\u57fa\u672c\u7684\u903b\u8f91\u56de\u5f52\u7f51\u7edc\u4f5c\u4e3a\u8d77\u70b9":13,"\u6211\u4eec\u5c06\u5728\u540e\u9762\u4ecb\u7ecd\u8bad\u7ec3\u548c\u9884\u6d4b\u7684\u6d41\u7a0b\u7684\u811a\u672c":13,"\u6211\u4eec\u5c06\u8bad\u7ec3\u7684\u547d\u4ee4\u884c\u4fdd\u5b58\u5728\u4e86":13,"\u6211\u4eec\u63a8\u8350\u4f7f\u7528docker\u955c\u50cf\u6765\u90e8\u7f72\u73af\u5883":8,"\u6211\u4eec\u63d0\u4f9b\u4e8612\u4e2a":9,"\u6211\u4eec\u63d0\u4f9b\u4e86\u4e00\u4e2a\u5de5\u5177\u7c7bdataproviderconvert":26,"\u6211\u4eec\u770b\u4e00\u4e0b\u5355\u53cc\u5c42\u5e8f\u5217\u7684\u4e0d\u540c\u6570\u636e\u7ec4\u7ec7\u5f62\u5f0f":1,"\u6211\u4eec\u770b\u4e00\u4e0b\u5355\u53cc\u5c42\u5e8f\u5217\u7684\u4e0d\u540c\u6570\u636e\u7ec4\u7ec7\u5f62\u5f0f\u548cdataprovid":1,"\u6211\u4eec\u770b\u4e00\u4e0b\u5355\u53cc\u5c42\u5e8f\u5217\u7684\u4e0d\u540cdataprovid":1,"\u6211\u4eec\u770b\u4e00\u4e0b\u5355\u5c42\u5e8f\u5217\u7684\u914d\u7f6e":1,"\u6211\u4eec\u770b\u4e00\u4e0b\u8bed\u4e49\u76f8\u540c\u7684\u53cc\u5c42\u5e8f\u5217\u914d\u7f6e":1,"\u6211\u4eec\u79f0\u4e4b\u4e3a\u4e00\u4e2a0\u5c42\u7684\u5e8f\u5217":0,"\u6211\u4eec\u8fdb\u5165\u5230\u8bad\u7ec3\u90e8\u5206":13,"\u6211\u4eec\u9009\u53d6\u5355\u53cc\u5c42\u5e8f\u5217\u914d\u7f6e\u4e2d\u7684\u4e0d\u540c\u90e8\u5206":1,"\u6211\u4eec\u91c7\u7528\u5355\u5c42lstm\u6a21\u578b":13,"\u6211\u4eec\u968f\u65f6\u603b\u7ed3\u4e86\u5404\u4e2a\u7f51\u7edc\u7684\u590d\u6742\u5ea6\u548c\u6548\u679c":13,"\u6216":1,"\u6216\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"\u6216\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":0,"\u6216\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":0,"\u6216\u4e00\u4e2a\u5411\u91cf":2,"\u6216\u5176\u4ed6":13,"\u6216\u5355\u5c42\u5e8f\u5217":0,"\u6216\u5355\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":0,"\u6216\u6700\u5927\u503c":0,"\u6216\u7b2c\u4e00\u4e2a":0,"\u6216\u7b2c\u4e00\u4e2a\u5143\u7d20":0,"\u6216\u8005":[0,9],"\u6216\u800510g\u8fd9\u6837\u7684\u5355\u4f4d":3,"\u6216\u8005\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":2,"\u6216\u8005\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":[0,2],"\u6216\u8005\u4f7f\u7528\u4e0b\u9762\u4e00\u6761\u547d\u4ee4\u5b89\u88c5":10,"\u6216\u8005\u5728python":23,"\u6216\u8005\u6570\u636e\u5e93\u8fde\u63a5\u5730\u5740\u7b49\u7b49":22,"\u6216\u8005\u662f\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":0,"\u6216\u8005\u8bbe\u7f6e\u4e3anone":22,"\u6216\u8005\u9700\u8981\u66f4\u9ad8\u7684\u6548\u7387":22,"\u6216\u8005\u9ad8\u6027\u80fd\u7684":9,"\u623f":1,"\u623f\u95f4":1,"\u6240\u4ee5":[23,26],"\u6240\u4ee5\u5728cpu\u7684\u8fd0\u7b97\u6027\u80fd\u4e0a\u5e76\u4e0d\u4f1a\u6709\u4e25\u91cd\u7684\u5f71\u54cd":9,"\u6240\u4ee5\u5982\u679c\u5bf9\u4e8e\u5185\u5b58\u6bd4\u8f83\u5c0f\u7684\u673a\u5668":23,"\u6240\u4ee5\u5982\u679c\u60f3\u8981\u5728\u540e\u53f0\u542f\u7528ssh":9,"\u6240\u4ee5\u5c06":23,"\u6240\u4ee5\u63a8\u8350\u4f7f\u7528\u663e\u5f0f\u6307\u5b9a\u8fd4\u56de\u503c\u548c\u6570\u636e\u5bf9\u5e94\u5173\u7cfb":23,"\u6240\u4ee5\u6700\u4f73\u5b9e\u8df5\u63a8\u8350\u4e0d\u8981\u5c06\u6bcf\u4e00\u4e2a\u6837\u672c\u90fd\u653e\u5165train":23,"\u6240\u4ee5\u7528\u4e8e\u9884\u6d4b\u7684\u914d\u7f6e\u6587\u4ef6\u8981\u505a\u76f8\u5e94\u7684\u4fee\u6539":26,"\u6240\u4ee5\u8f93\u51fa\u7684value\u5305\u542b\u4e24\u4e2a\u5411\u91cf":26,"\u6240\u4ee5gpu\u5728\u8fd0\u7b97\u6027\u80fd\u4e0a\u4e5f\u4e0d\u4f1a\u6709\u4e25\u91cd\u7684\u5f71\u54cd":9,"\u6240\u4ee5init_hook\u5c3d\u91cf\u4f7f\u7528":23,"\u6240\u6709\u5b57\u7b26\u90fd\u5c06\u8f6c\u6362\u4e3a\u8fde\u7eed\u6574\u6570\u8868\u793a\u7684id\u4f20\u7ed9\u6a21\u578b":13,"\u6240\u6709\u6587\u4ef6\u5217\u8868":23,"\u6240\u6709\u7684":4,"\u6240\u6709\u7684\u4e0b\u8f7d\u7ebf\u7a0b\u5173\u95ed\u65f6":3,"\u6240\u6709\u914d\u7f6e\u5728":13,"\u6240\u8c13\u65f6\u95f4\u6b65\u4fe1\u606f":23,"\u624d\u4f1a\u91ca\u653e\u8be5\u6bb5\u5185\u5b58":23,"\u624d\u4f1astop":23,"\u624d\u80fd\u4fdd\u8bc1\u548c\u5355\u5c42\u5e8f\u5217\u7684\u914d\u7f6e\u4e2d":1,"\u6253\u5370\u7684\u65e5\u5fd7\u53d8\u591a":4,"\u6267\u884c":3,"\u6267\u884c\u5982\u4e0b\u547d\u4ee4\u5373\u53ef\u4ee5\u5173\u95ed\u8fd9\u4e2acontain":9,"\u6267\u884c\u65b9\u6cd5\u5982\u4e0b":9,"\u62a5\u9519":10,"\u62c6\u89e3":2,"\u62fc\u63a5\u6210\u4e00\u4e2a\u65b0\u7684\u5411\u91cf\u8868\u793a":13,"\u6307\u4ee4\u96c6":9,"\u6307\u5411\u4e00\u4e2alayer":2,"\u6307\u5b9a":2,"\u6307\u5b9a\u521d\u59cb\u5316\u6a21\u578b\u8def\u5f84":13,"\u6307\u5b9a\u751f\u6210\u6570\u636e\u7684\u51fd\u6570":13,"\u6307\u5b9a\u7684\u8f93\u5165\u4e0d\u4f1a\u88ab":2,"\u6307\u5b9a\u8bad\u7ec3":13,"\u6307\u5b9abatch":13,"\u6307\u5b9aoutputs\u8f93\u51fa\u6982\u7387\u5c42":13,"\u633a":1,"\u633a\u597d":1,"\u6362":1,"\u6389\u7f16\u8bd1\u76ee\u5f55\u540e":4,"\u6392\u6210\u4e00\u5217\u7684\u591a\u4e2a\u5143\u7d20":0,"\u63a5\u4e0b\u6765\u4f7f\u7528":26,"\u63a5\u53e3\u4f7f\u7528\u591a\u7ebf\u7a0b\u8bfb\u53d6\u6570\u636e":23,"\u63a5\u7740":1,"\u63a8\u8350":1,"\u63a8\u8350\u4f7f\u7528\u5c06\u672c\u5730\u7f51\u5361":9,"\u63a8\u8350\u4f7f\u7528\u6700\u65b0\u7248\u672c\u7684cudnn":4,"\u63a8\u8350\u6e05\u7406":4,"\u63a8\u8350\u76f4\u63a5\u653e\u7f6e\u5230\u8bad\u7ec3\u76ee\u5f55":22,"\u63a8\u8350\u8bbe\u7f6e":23,"\u63cf\u8ff0":4,"\u63cf\u8ff0\u4e86docker":3,"\u63d0\u4f9b\u6269\u5c55\u7684\u957f\u5ea6\u4fe1\u606f":0,"\u653e\u5fc3":1,"\u6548\u679c\u4e00\u81f4":23,"\u6548\u679c\u603b\u7ed3":13,"\u6559\u7a0b":13,"\u6570":2,"\u6570\u5fc5\u987b\u4e25\u683c\u76f8\u7b49":2,"\u6570\u636e":23,"\u6570\u636e\u4f20\u8f93\u65e0\u9700label\u6570\u636e":13,"\u6570\u636e\u5904\u7406python\u6587\u4ef6\u540d":13,"\u6570\u636e\u5982\u4f55\u5b58\u50a8\u7b49\u7b49":23,"\u6570\u636e\u63d0\u4f9b":22,"\u6570\u636e\u6587\u4ef6\u5b58\u653e\u5728\u672c\u5730\u78c1\u76d8\u4e2d":22,"\u6570\u636e\u662f\u7ed9\u4e00\u6bb5\u82f1\u6587\u6587\u672c":23,"\u6570\u636e\u683c\u5f0f\u548c\u8be6\u7ec6\u6587\u6863\u8bf7\u53c2\u8003":13,"\u6570\u636e\u8f93\u5165":2,"\u6574\u4f53":1,"\u6574\u6d01":1,"\u6587\u4ef6":23,"\u6587\u4ef6\u4e2d":13,"\u6587\u672c\u4e2d\u7684\u5355\u8bcd\u7528\u7a7a\u683c\u5206\u9694":13,"\u6587\u672c\u4fe1\u606f\u5c31\u662f\u4e00\u4e2a\u5e8f\u5217":23,"\u6587\u672c\u5206\u7c7b\u95ee\u9898":13,"\u6587\u672c\u5377\u79ef\u5206\u4e3a\u4e09\u4e2a\u6b65\u9aa4":13,"\u6587\u672c\u751f\u6210":12,"\u65b0":1,"\u65b0\u5199layer":14,"\u65b9\u4fbf":1,"\u65b9\u4fbf\u8c03\u8bd5\u4f7f\u7528":16,"\u65b9\u4fbf\u90e8\u7f72\u5206\u53d1":16,"\u65c1\u8fb9":1,"\u65e0":1,"\u65e0\u6cd5\u76f4\u63a5\u4f7f\u7528":1,"\u65e0\u9700label\u76f8\u5173\u7684\u5c42":13,"\u65e9\u9910":1,"\u65f6":0,"\u65f6\u5019":1,"\u65f6\u5e8f\u6a21\u578b\u5373\u4e3arnn\u6a21\u578b":13,"\u65f6\u5e8f\u6a21\u578b\u5747\u4f7f\u7528\u8be5\u811a\u672c":13,"\u662f":[1,4],"\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"\u662f\u4e00\u4e2a\u53cc\u5c42\u7684\u5e8f\u5217":0,"\u662f\u4e00\u4e2abatch":23,"\u662f\u4e00\u4e2apython\u7684":23,"\u662f\u4e00\u4e2aswig\u5c01\u88c5\u7684paddlepaddle\u5305":9,"\u662f\u4e00\u4e2aunbound":2,"\u662f\u4e00\u79cd\u4efb\u610f\u590d\u6742\u7684rnn\u5355\u5143":2,"\u662f\u4e0d\u662f\u5f88\u7b80\u5355\u5462":23,"\u662f\u4e2adataprovider\u662f\u4e0d\u662f\u8981\u505ashuffl":23,"\u662f\u4ec0\u4e48\u4e5f\u6ca1\u5173\u7cfb":23,"\u662f\u4ece\u8bad\u7ec3\u914d\u7f6e\u4f20\u5165\u7684dict\u5bf9\u8c61":23,"\u662f\u51e0\u4e4e\u4e0d\u5360\u5185\u5b58\u7684":23,"\u662f\u521d\u59cb\u5316\u65f6\u8c03\u7528\u7684\u51fd\u6570":23,"\u662f\u540c\u4e00\u4e2a\u5bf9\u8c61":23,"\u662f\u5426\u4ee5\u9006\u5e8f\u5904\u7406\u8f93\u5165\u5e8f\u5217":2,"\u662f\u5426\u4f7f\u7528\u53cc\u7cbe\u5ea6\u6d6e\u70b9\u6570":4,"\u662f\u5426\u4f7f\u7528\u8fd0\u884c\u65f6\u52a8\u6001\u52a0\u8f7dcuda\u52a8\u6001\u5e93":4,"\u662f\u5426\u4f7f\u7528gflags":4,"\u662f\u5426\u4f7f\u7528glog":4,"\u662f\u5426\u5185\u5d4cpython\u89e3\u91ca\u5668":4,"\u662f\u5426\u5bfb\u627e\u5230cuda\u5de5\u5177\u94fe":4,"\u662f\u5426\u5f00\u542f\u5355\u5143\u6d4b\u8bd5":4,"\u662f\u5426\u5f00\u542f\u8ba1\u65f6\u529f\u80fd\u5f00\u542f\u8ba1\u65f6\u529f\u80fd\u4f1a\u5bfc\u81f4\u8fd0\u884c\u7565\u6162":4,"\u662f\u5426\u5f00\u542fgpu\u529f\u80fd":3,"\u662f\u5426\u5f00\u542frdma\u652f\u6301":4,"\u662f\u5426\u7f16\u8bd1\u4e2d\u6587\u6587\u6863":4,"\u662f\u5426\u7f16\u8bd1\u542b\u6709avx\u6307\u4ee4\u96c6\u7684paddlepaddle\u4e8c\u8fdb\u5236":4,"\u662f\u5426\u7f16\u8bd1\u65f6\u8fdb\u884c\u4ee3\u7801\u98ce\u683c\u68c0\u67e5":4,"\u662f\u5426\u7f16\u8bd1\u82f1\u6587\u6587\u6863":4,"\u662f\u5426\u7f16\u8bd1gpu\u652f\u6301":4,"\u662f\u5426\u7f16\u8bd1python\u7684swig\u63a5\u53e3":4,"\u662f\u5728\u8fd0\u884c\u65f6\u6267\u884c\u7684":23,"\u662f\u5f85\u6269\u5c55\u7684\u6570\u636e":0,"\u662f\u6570\u636e\u7f13\u5b58\u7684\u7b56\u7565":23,"\u662f\u6570\u636e\u8f93\u5165\u683c\u5f0f":23,"\u662f\u8bbe\u7f6e\u8fd9\u4e2adataprovider\u8fd4\u56de\u4ec0\u4e48\u6837\u7684\u6570\u636e":23,"\u662f\u8bbe\u7f6edataprovider\u5728\u5185\u5b58\u4e2d\u6682\u5b58\u7684\u6570\u636e\u6761\u6570":23,"\u662f\u8bbe\u7f6edataprovider\u5728\u5185\u5b58\u4e2d\u6700\u5c0f\u6682\u5b58\u7684\u6570\u636e\u6761\u6570":23,"\u662fdecoder\u7684\u6570\u636e\u8f93\u5165":2,"\u662fpaddlepaddle\u652f\u6301\u7684\u4e00\u79cd\u4efb\u610f\u590d\u6742\u7684rnn\u5355\u5143":2,"\u662fpaddlepaddle\u8d1f\u8d23\u63d0\u4f9b\u6570\u636e\u7684\u6a21\u5757":22,"\u662fpython\u7684\u4e00\u4e2a\u5173\u952e\u8bcd":23,"\u663e":13,"\u665a":1,"\u666e\u901a\u7528\u6237\u8bf7\u8d70\u5b89\u88c5\u6d41\u7a0b":8,"\u66f4\u597d\u5730\u5b8c\u6210\u4e00\u4e9b\u590d\u6742\u7684\u8bed\u8a00\u7406\u89e3\u4efb\u52a1":2,"\u66f4\u8be6\u7ec6\u7528\u4f8b\u8bf7\u53c2\u8003\u6587\u6863":13,"\u66f4\u8be6\u7ec6\u7684\u4ecb\u7ecd\u8bf7\u53c2\u8003\u5404\u4e2a\u547d\u4ee4\u7684\u547d\u4ee4\u884c\u53c2\u6570\u6587\u6863":16,"\u66f4\u8be6\u7ec6\u7684\u7f51\u7edc\u914d\u7f6e":13,"\u66f4\u8fdb\u4e00\u6b65":2,"\u66ff\u6211\u4eec\u5b8c\u6210\u4e86\u539f\u59cb\u8f93\u5165\u6570\u636e\u7684\u62c6\u5206":2,"\u6700":1,"\u6700\u4f4e\u7ebf\u7a0b\u7684\u4e0b\u8f7d\u901f\u5ea6":3,"\u6700\u540e":1,"\u6700\u540e\u4e00\u4e2a":0,"\u6700\u540e\u4f7f\u7528":26,"\u6700\u7ec8\u5b9e\u73b0\u4e00\u4e2a\u5c42\u6b21\u5316\u7684\u590d\u6742rnn":2,"\u6700\u7ec8\u7684\u8f93\u51fa\u7ed3\u679c":2,"\u6708\u6e56":1,"\u6709":1,"\u6709100\u4e2a\u8bad\u7ec3\u6587\u4ef6":23,"\u6709\u4e24\u53e5":1,"\u6709\u503c\u7684\u4f4d\u7f6e\u53ea\u80fd\u53d61":23,"\u6709\u503c\u7684\u90e8\u5206\u53ef\u4ee5\u662f\u4efb\u4f55\u6d6e\u70b9\u6570":23,"\u6709\u90e8\u5206\u53c2\u6570\u662fpaddle\u81ea\u52a8\u751f\u6210\u7684":23,"\u670d\u52a1":1,"\u670d\u52a1\u5458":1,"\u672c\u6765":1,"\u672c\u8282\u6211\u4eec\u5c06\u4e13\u6ce8\u4e8e\u7f51\u7edc\u7ed3\u6784\u7684\u4ecb\u7ecd":13,"\u6765":1,"\u6765\u5b89\u88c5":10,"\u6765\u5bf9\u6bd4\u5206\u6790\u4e24\u8005\u8bed\u4e49\u76f8\u540c\u7684\u539f\u56e0":1,"\u6765\u5f15\u7528\u8fd9\u4e2aimag":9,"\u6765\u63a5\u53d7\u4e0d\u4f7f\u7528\u7684":23,"\u6765\u786e\u5b9a\u5bf9\u5e94\u5173\u7cfb":23,"\u6765\u81ea\u5b9a\u4e49\u4f20\u6570\u636e\u7684\u8fc7\u7a0b":22,"\u6765\u8bf4\u660e\u7b80\u5355\u7684pydataprovider\u5982\u4f55\u4f7f\u7528":23,"\u6765\u8fdb\u884c\u8bad\u7ec3":9,"\u6765\u914d\u7f6ecudnn\u7684\u5b89\u88c5\u8def\u5f84":4,"\u676f\u5b50":1,"\u6784\u6210\u4e86\u8f93\u51fa\u53cc\u5c42\u5e8f\u5217\u7684\u7b2ci\u4e2asubseq":0,"\u6784\u9020gradientmachin":26,"\u6790\u597d\u7684\u914d\u7f6e\u521b\u5efa\u795e\u7ecf\u7f51\u7edc":26,"\u67e5\u770b\u5b89\u88c5\u540e\u7684paddl":10,"\u6807\u7b7e\u662f0":23,"\u6837\u4f8b\u6570\u636e\u4e3a":23,"\u6837\u4f8b\u6570\u636e\u5982\u4e0b":23,"\u6837\u672c":23,"\u6837\u672c\u95f4\u7528\u7a7a\u884c\u5206\u5f00":1,"\u6839\u636e\u4e0a\u4e00\u6b65\u89e3":26,"\u6839\u636e\u6a21\u578b\u914d\u7f6e\u6587\u4ef6\u4e2d":23,"\u683c\u5f0f\u5982\u4e0b":13,"\u68d2":13,"\u697c\u5c42":1,"\u6a21\u578b\u5b58\u50a8\u8def\u5f84":13,"\u6a21\u578b\u8bad\u7ec3\u4f1a\u770b\u5230\u8fd9\u6837\u7684\u65e5\u5fd7":13,"\u6a21\u578b\u914d\u7f6e":14,"\u6a2a\u5411\u5305\u62ec\u4e09\u4e2a\u7248\u672c":9,"\u6b21":1,"\u6b63\u5e38\u7684docker":9,"\u6b63\u6837\u672c":13,"\u6b64\u5904\u90fd\u4e3a2":1,"\u6bb5\u843d\u53ef\u4ee5\u770b\u4f5c\u662f\u4e00\u4e2a\u5d4c\u5957\u7684\u53cc\u5c42\u7684\u5e8f\u5217":2,"\u6bcf\u4e00\u4e2a\u4efb\u52a1\u6d41\u7a0b\u90fd\u53ef\u4ee5\u5206\u4e3a\u5982\u4e0b5\u4e2a\u57fa\u7840\u90e8\u5206":13,"\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65":1,"\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u90fd\u7528\u4e86\u4e0a\u4e00\u4e2a\u65f6\u95f4\u6b65\u7684\u8f93\u51fa\u7ed3\u679c":1,"\u6bcf\u4e00\u6761\u8bad\u7ec3\u6570\u636e\u90fd\u662f\u4e00\u4e2a\u6587\u4ef6":23,"\u6bcf\u4e00\u884c":23,"\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2a0\u5c42\u5e8f\u5217":0,"\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"\u6bcf\u4e2a\u5355\u5c42rnn":2,"\u6bcf\u4e2a\u5c42\u90fd\u6709\u4e00\u4e2a\u6216\u591a\u4e2ainput":13,"\u6bcf\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185\u7684\u8fd0\u7b97\u662f\u72ec\u7acb\u7684":2,"\u6bcf\u4e2a\u6837\u672c\u7531\u4e24\u90e8\u5206\u7ec4\u6210":1,"\u6bcf\u4e2a\u6837\u672c\u7684\u5b50\u53e5\u6570\u5206\u522b\u4e3a2":1,"\u6bcf\u4e2a\u72b6\u6001":2,"\u6bcf\u4e2agenerator\u5728\u6ca1\u6709\u8c03\u7528\u7684\u65f6\u5019":23,"\u6bcf\u4e2apass\u7684\u7b2c0\u4e2abatch\u5230\u5f53\u524dbatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747\u5206\u7c7b\u9519\u8bef\u7387":13,"\u6bcf\u4e2apass\u7684\u7b2c0\u4e2abatch\u5230\u5f53\u524dbatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747cost":13,"\u6bcf\u4e2asubseq\u7684\u6700\u540e\u4e00\u4e2a\u5143\u7d20\u5c31\u7b49\u4e8e\u5355\u5c42\u5e8f\u5217\u7684\u6700\u540e\u4e00\u4e2a\u5143\u7d20":1,"\u6bcf\u6b21\u90fd\u4f1a\u4ecepython\u7aef\u8bfb\u53d6\u6570\u636e":23,"\u6bcf\u884c\u4fdd\u5b58\u4e00\u6761\u6837\u672c":13,"\u6bcf\u9694\u591a\u5c11batch\u6253\u5370\u4e00\u6b21\u65e5\u5fd7":13,"\u6bd4\u5982\u901a\u8fc7\u7528\u6237\u5bf9\u7535\u5b50\u5546\u52a1\u7f51\u7ad9\u8bc4\u8bba":13,"\u6bd4\u8f83\u53ef\u80fd\u7684\u547d\u4ee4\u5982\u4e0b":10,"\u6c34\u6e29":1,"\u6c49\u5ead":1,"\u6ca1":1,"\u6ca1\u6709\u4f5c\u7528":23,"\u6ca1\u6709\u5b89\u88c5":10,"\u6ca1\u6709\u8bbe\u7f6e":10,"\u6ce8\u610f":[1,3,23],"\u6cf3\u6c60":1,"\u6d41":1,"\u6d41\u7a0b\u5982\u4e0b":13,"\u6d44":1,"\u6d4b\u8bd5\u6570\u636e":13,"\u6d4b\u8bd5\u7684\u65f6\u5019\u9ed8\u8ba4\u4e0dshuffl":23,"\u6d4b\u8bd5\u811a\u672c\u5982\u4e0b":13,"\u6e29\u99a8":1,"\u6e90\u7801":13,"\u6e90\u7801\u6839\u76ee\u5f55":3,"\u6fc0\u6d3b\u51fd\u6570\u7c7b\u578b":13,"\u70ed\u60c5":1,"\u7136\u540e\u4ea4\u7ed9step\u51fd\u6570":2,"\u7136\u540e\u6267\u884c\u5982\u4e0b":10,"\u7136\u540e\u8fd0\u884c\u8fd9\u4e2acontainer\u5373\u53ef":9,"\u7248\u672c":10,"\u751f\u6210\u5404\u4e2a\u5e73\u53f0\u7684makefil":4,"\u75280\u548c1\u8868\u793a":23,"\u7528\u4e86\u4e24\u4e2a\u6708\u4e4b\u540e\u8fd9\u4e2a\u663e\u793a\u5668\u5c4f\u5e55\u788e\u4e86":13,"\u7528\u4e8e\u4e0d\u652f\u6301avx\u6307\u4ee4\u96c6\u7684cpu":10,"\u7528\u6237\u4e5f\u53ef\u4ee5\u5728c":22,"\u7528\u6237\u4e5f\u53ef\u4ee5\u663e\u5f0f\u6307\u5b9a\u8fd4\u56de\u7684\u6570\u636e\u5bf9\u5e94\u5173\u7cfb":23,"\u7528\u6237\u53ea\u9700\u5b9a\u4e49rnn\u5728\u4e00\u4e2a\u65f6\u95f4\u6b65\u5185\u5b8c\u6210\u7684\u8ba1\u7b97":2,"\u7528\u6237\u53ef\u4ee5\u4f7f\u7528python\u7684":22,"\u7528\u6237\u53ef\u4ee5\u6839\u636e\u8bad\u7ec3log\u9009\u62e9test\u7ed3\u679c\u6700\u597d\u7684\u6a21\u578b\u6765\u9884\u6d4b":13,"\u7528\u6237\u53ef\u4ee5\u9009\u62e9\u5bf9\u5e94\u7248\u672c\u7684docker":9,"\u7528\u6237\u540d\u4e3a":9,"\u7528\u6237\u5728dataprovider\u4e2d\u9700\u8981\u5b9e\u73b0\u5982\u4f55\u8bbf\u95ee\u5176\u4e2d\u6bcf\u4e00\u4e2a\u6587\u4ef6":22,"\u7528\u6237\u5b9a\u4e49\u7684\u53c2\u6570\u4f7f\u7528args\u5728\u8bad\u7ec3\u914d\u7f6e\u4e2d\u8bbe\u7f6e":23,"\u7528\u6237\u63a5\u53e3":14,"\u7528\u6237\u9700\u8981\u5148\u5c06paddlepaddle\u5b89\u88c5\u5305\u4e0b\u8f7d\u5230\u672c\u5730":10,"\u7528\u6765\u505a\u9884\u6d4b\u548c\u7b80\u5355\u7684\u5b9a\u5236\u5316":9,"\u7528\u8fc7\u4e00\u6b21\u7684\u65f6\u5019":23,"\u7531":2,"\u7531\u4e8e\u5916\u5c42\u6bcf\u4e2a\u65f6\u95f4\u6b65\u8fd4\u56de\u7684\u662f\u4e00\u4e2a\u5b50\u53e5":1,"\u7531\u4e8e\u5916\u5c42memory\u6ca1\u6709\u4efb\u4f55seq\u4fe1\u606f":1,"\u7531\u4e8e\u6570\u636e\u662f\u4e24\u6761":26,"\u7531\u4e8e\u8fd9\u4e2a\u5916\u5c42group\u91cc\u9762\u6ca1\u6709memori":1,"\u7531\u4e8edocker\u662f\u57fa\u4e8e\u5bb9\u5668\u7684\u8f7b\u91cf\u5316\u865a\u62df\u65b9\u6848":9,"\u7531\u4e8epaddlepaddle\u7684docker\u955c\u50cf\u5e76\u4e0d\u5305\u542b\u4efb\u4f55\u9884\u5b9a\u4e49\u7684\u8fd0\u884c\u547d\u4ee4":9,"\u7531\u4e8estep":2,"\u7531\u6613\u5230\u96be\u5c55\u793a4\u79cd\u4e0d\u540c\u7684\u7f51\u7edc\u914d\u7f6e":13,"\u7531\u8bcd\u8bed\u6784\u6210\u7684\u53e5\u5b50":0,"\u7535\u8111":1,"\u7684":[1,13],"\u7684\u4e00\u4e2a\u7b80\u5355\u8c03\u7528\u5982\u4e0b":2,"\u7684\u540d\u5b57":23,"\u7684\u5b89\u88c5\u6587\u6863":9,"\u7684\u5e73\u5747\u503c":0,"\u7684\u60c5\u51b5\u4e0b\u8d8a\u5927\u8d8a\u597d":23,"\u7684\u6570\u76ee\u4e00\u81f4":0,"\u7684\u6587\u6863":23,"\u7684\u65f6\u5019\u5982\u679c\u62a5\u4e00\u4e9b\u4f9d\u8d56\u672a\u627e\u5230\u7684\u9519\u8bef\u662f\u6b63\u5e38\u7684":10,"\u7684\u662f":23,"\u7684\u673a\u5668\u4e0a\u8fdb\u884c":3,"\u7684\u6838\u5fc3\u662f\u8bbe\u8ba1step\u51fd\u6570\u7684\u8ba1\u7b97\u903b\u8f91":2,"\u7684\u6bb5\u843d\u5b9a\u4e49\u4e3a\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":2,"\u7684\u72b6\u6001":2,"\u7684\u7f51\u6865\u6765\u8fdb\u884c\u7f51\u7edc\u901a\u4fe1":9,"\u7684\u8f93\u5165":2,"\u7684\u9519\u8bef":1,"\u7684demo\u5b66\u4e60\u5982\u4f55\u8fdb\u884c\u591a\u673a\u8bad\u7ec3":13,"\u7684docker\u53ef\u80fd\u7f3a\u4e4f":3,"\u7684matrix":26,"\u7684python\u5305\u662fpaddlepaddle\u7684\u8bad\u7ec3\u4e3b\u8981\u7a0b\u5e8f":9,"\u7684python\u5305\u6765\u505a\u914d\u7f6e\u6587\u4ef6\u89e3\u6790\u7b49\u5de5\u4f5c":9,"\u7684python\u9884\u6d4b\u8fc7\u7a0b":13,"\u76ee\u524d":2,"\u76ee\u524d\u652f\u6301\u4e24\u79cd":0,"\u76ee\u524d\u8fd8\u672a\u652f\u6301":2,"\u76ee\u5f55":13,"\u76ee\u5f55\u4e0b":3,"\u76f4\u63a5\u52a0\u4e86\u4e00\u5c42group":1,"\u76f4\u63a5\u63d0\u53d6\u51fa\u795e\u7ecf\u7f51\u7edcoutput\u5c42\u7684\u8f93\u51fa\u7ed3\u679c":26,"\u76f8\u5173\u547d\u4ee4\u4e3a":9,"\u76f8\u5173\u7684\u6982":23,"\u76f8\u5bf9":1,"\u76f8\u5bf9\u4e8epaddlepaddle\u7a0b\u5e8f\u8fd0\u884c\u65f6\u7684\u8def\u5f84":22,"\u76f8\u5f53":1,"\u77e5\u9053\u5982\u4f55\u4ece":23,"\u793a":13,"\u79bb":1,"\u79f0\u4e4b\u4e3a\u53cc\u5c42\u5e8f\u5217\u7684\u4e00\u4e2a\u5b50\u5e8f\u5217":0,"\u7a0b\u5e8f\u6216\u8005\u81ea\u5b9a\u4e49\u4e00\u4e2a\u542b\u6709\u542f\u52a8\u811a\u672c\u7684imag":9,"\u7a97\u6237":1,"\u7aef\u81ea\u5b9a\u4e49\u4e00\u4e2a":22,"\u7b2c":1,"\u7b2c\u4e00\u4e2a\u53c2\u6570\u662f":23,"\u7b2c\u4e00\u4e2alast":1,"\u7b2c\u4e00\u4e2apass\u4f1a\u4ecepython\u7aef\u8bfb\u53d6\u6570\u636e":23,"\u7b2c\u4e00\u5929":1,"\u7b2c\u4e00\u6bb5\u6570\u636e\u4e3a\u8fd9\u5f20\u56fe\u7247\u7684label":23,"\u7b2c\u4e8c\u4e2a\u53c2\u6570\u662ffilenam":23,"\u7b2c\u4e8c\u6bb5\u6570\u636e\u4e3a\u8fd9\u4e2a\u56fe\u7247\u7684\u50cf\u7d20\u503c":23,"\u7b80\u5355\u4f18\u5316":3,"\u7b80\u5355\u7684\u4f7f\u7528":22,"\u7b80\u5355\u7684\u4f7f\u7528\u573a\u666f":22,"\u7b80\u5355\u7684\u4f7f\u7528\u6837\u4f8b\u4e3a":3,"\u7b80\u5355\u7684\u542b\u6709ssh\u7684dockerfile\u5982\u4e0b":9,"\u7b80\u5355\u7684pydataprovider\u6837\u4f8b\u5c31\u8bf4\u660e\u5b8c\u6bd5\u4e86":23,"\u7b80\u76f4":1,"\u7c7b\u522bid":13,"\u7c7b\u522bid\u7684\u6570\u636e\u7c7b\u578b":13,"\u7c7b\u578b\u53ef\u4ee5\u662fpaddlepaddle\u652f\u6301\u7684\u4efb\u610f\u8f93\u5165\u6570\u636e\u7c7b\u578b":0,"\u7c7b\u578b\u6765\u8bbe\u7f6e":23,"\u7eb5\u5411\u5305\u62ec\u56db\u4e2a\u7248\u672c":9,"\u7ec3":16,"\u7ed3\u4e0a\u8ff0\u7f51\u7edc\u7ed3\u6784\u5728amazon":13,"\u7ed9":1,"\u7ed9\u5b9aencoder\u8f93\u51fa\u548c\u5f53\u524d\u8bcd":2,"\u7ee7\u7eed\u8bad\u7ec3":23,"\u7ef4\u5ea6\u4e3aword":13,"\u7ef4\u5ea6\u662f\u7c7b\u522b\u4e2a\u6570":13,"\u7ef4\u5ea6\u662f\u8bcd\u5178\u5927\u5c0f":13,"\u7f13\u5b58\u8bad\u7ec3\u6570\u636e\u5230\u5185\u5b58":23,"\u7f16\u8bd1\u53c2\u6570\u9009\u9879\u6587\u4ef6":21,"\u7f16\u8bd1\u73af\u5883\u548c\u6e90\u4ee3\u7801":9,"\u7f16\u8bd1\u9009\u9879":4,"\u7f16\u8bd1\u9009\u9879\u4e3b\u8981\u63a8\u8350\u9ad8\u7ea7\u7528\u6237\u67e5\u770b":8,"\u7f16\u8bd1\u9009\u9879\u5217\u8868\u5982\u4e0b":4,"\u7f16\u8bd1paddlepaddle\u7684gpu\u7248\u672c\u5e76\u4e0d\u9700\u8981\u4e00\u5b9a\u5728\u5177\u6709gpu":3,"\u7f51\u7edc\u540d\u79f0":13,"\u7f51\u7edc\u914d\u7f6e":13,"\u7f6e\u8fd9\u4e9b\u53d8\u91cf":4,"\u800c":9,"\u800c\u4e09\u79cd\u5e8f\u5217\u6a21\u5f0f\u4e3a":23,"\u800c\u4e0d\u4f7f\u7528docker":9,"\u800c\u4e0d\u7528\u5173\u5fc3\u6570\u636e\u5982\u4f55\u4f20\u8f93\u7ed9paddlepaddl":23,"\u800c\u4e14\u9884\u6d4b\u7f51\u7edc\u901a\u5e38\u76f4\u63a5\u8f93\u51fa\u6700\u540e\u4e00\u5c42\u7684\u7ed3\u679c\u800c\u4e0d\u662f\u50cf\u8bad\u7ec3\u65f6\u4e00\u6837\u4ee5cost":26,"\u800c\u5728":[4,23],"\u800c\u5982\u679c\u6309\u987a\u5e8f\u8c03\u7528\u8fd9\u4e9bgenerator\u5c31\u4e0d\u4f1a\u51fa\u73b0\u8fd9\u4e2a\u95ee\u9898":23,"\u800c\u662f\u5c06\u6837\u672c\u7684\u5730\u5740\u653e\u5165\u53e6\u4e00\u4e2a\u6587\u672c":23,"\u800c\u6ca1\u6709\u6d4b\u8bd5\u6570\u636e":23,"\u800c\u7279\u5f81\u5373\u4e3a":23,"\u800c\u8fd9\u4e2a\u4e00\u822c\u8bf4\u660epaddlepaddle\u5df2\u7ecf\u5b89\u88c5\u5b8c\u6bd5\u4e86":10,"\u800c\u8fd9\u4e2a\u53d8\u91cf\u63a8\u8350\u5927\u4e8e\u8bad\u7ec3\u7684batch":23,"\u800c\u8fd9\u4e2acontext\u53ef\u80fd\u4f1a\u975e\u5e38":23,"\u800c\u975e\u9759\u6001\u52a0\u8f7dcuda\u52a8\u6001\u5e93":4,"\u800cexpand":1,"\u800cgpu\u7684\u9a71\u52a8\u548c\u8bbe\u5907\u5168\u90e8\u6620\u5c04\u5230\u4e86\u5bb9\u5668\u5185":9,"\u800cpaddlepaddle\u8fdb\u7a0b\u5e2e\u52a9\u7528\u6237\u505a\u4e86":23,"\u800crnn\u662f\u6700\u6d41\u884c\u7684\u9009\u62e9":2,"\u80fd\u591f\u5904\u7406\u53cc\u5c42\u5e8f\u5217":2,"\u80fd\u591f\u5bf9\u53cc\u5411\u5e8f\u5217\u8fdb\u884c\u5904\u7406\u7684\u6709":2,"\u80fd\u591f\u8bb0\u5f55\u4e0a\u4e00\u4e2asubseq":2,"\u811a\u672c":9,"\u811a\u672c\u53ef\u4ee5\u542f\u52a8paddlepaddle\u7684\u8bad\u7ec3\u8fdb\u7a0b\u548cpserv":9,"\u811a\u672c\u548c":9,"\u811a\u672c\u7c7b\u4f3c\u4e8e":9,"\u81ea\u52a8\u5b8c\u6210\u8fd9\u4e00\u8fc7\u7a0b":2,"\u81ea\u5b9a\u4e49\u4e00\u4e2adataprovid":22,"\u81f3\u5c11\u5177\u67093":9,"\u81f3\u6b64":[9,23],"\u8212\u9002":1,"\u82e5\u5e72\u4e2a\u53e5\u5b50\u6784\u6210\u4e00\u4e2a\u6bb5\u843d":0,"\u82e5\u8f93\u51fa\u662f\u5355\u5c42\u5e8f\u5217":0,"\u82e5\u8f93\u51fa\u662f\u53cc\u5c42\u5e8f\u5217":0,"\u83b7\u53d6\u5229\u7528one":13,"\u83b7\u53d6\u6bcf\u4e2a\u5355\u8bcd\u5de6\u53f3\u5404k\u4e2a\u8fd1\u90bb":13,"\u83b7\u53d6\u8be5\u6761\u6837\u672c\u7c7b\u522bid":13,"\u8868\u793a\u5c06\u5916\u5c42\u7684outer":1,"\u8868\u793a\u6574\u6570\u6807\u7b7e":23,"\u8868\u793a\u662f\u5426\u5141\u8bb8paddle\u6682\u5b58\u7565\u5fae\u591a\u4f59pool_size\u7684\u6570\u636e":23,"\u8868\u793a\u7a00\u758f\u7684\u5411\u91cf":23,"\u8868\u793a\u7a00\u758f\u7684\u96f6\u4e00\u5411\u91cf":23,"\u8868\u793a\u7a20\u5bc6\u7684\u6d6e\u70b9\u6570\u5411\u91cf":23,"\u8868\u793a\u8fc7\u4e8620\u4e2abatch":13,"\u8868\u793a\u8fc7\u4e862560\u4e2a\u6837\u672c":13,"\u8868\u793a\u8fd9\u4e2adataprovider\u662f\u8bad\u7ec3\u7528\u7684dataprovider\u6216\u8005\u6d4b\u8bd5\u7528\u7684":23,"\u8868\u793asubseq\u95f4\u4e0d\u5b58\u5728\u8054\u7cfb":1,"\u88ab\u6269\u5c55\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"\u8981\u6c42\u5355\u5c42\u5e8f\u5217\u542b\u6709\u5143\u7d20\u7684\u6570\u76ee":0,"\u8981\u751f\u6210\u7684\u76ee\u6807\u5e8f\u5217":2,"\u89c1":1,"\u89e3\u51b3\u529e\u6cd5\u662f\u5c06cuda":10,"\u89e3\u51b3\u65b9\u6cd5\u5f88\u7b80\u5355":10,"\u89e3\u6790\u8bad\u7ec3\u65f6\u7684\u914d\u7f6e\u6587\u4ef6":26,"\u89e3\u91ca":13,"\u8ba9\u795e\u7ecf\u7f51\u7edc\u53ef\u4ee5\u8fdb\u884c\u8bad\u7ec3":22,"\u8bad\u7ec3":9,"\u8bad\u7ec3\u6570\u636e\u975e\u5e38\u591a\u7684\u60c5\u51b5\u4e0b":23,"\u8bad\u7ec3\u6587\u4ef6\u5217\u8868":22,"\u8bad\u7ec3\u65f6\u6240\u9700\u8bbe\u7f6e\u7684\u4e3b\u8981\u53c2\u6570\u5982\u4e0b":13,"\u8bad\u7ec3\u7684\u65f6\u5019\u9ed8\u8ba4shuffl":23,"\u8bad\u7ec3\u811a\u672c":13,"\u8bad\u7ec3\u811a\u672c\u5728":13,"\u8bad\u7ec3\u8f6e\u6b21":13,"\u8bb2\u6570\u636e\u53d1\u9001\u7ed9paddlepaddl":23,"\u8bb2\u89e3\u5982\u4f55\u4f7f\u7528\u53cc\u5c42rnn":1,"\u8bbe\u7f6e\u4e0b\u5217\u7f16\u8bd1\u9009\u9879\u65f6":4,"\u8bbe\u7f6e\u6210":23,"\u8bbe\u7f6e\u6210\u4e86\u5e8f\u5217":23,"\u8bbe\u7f6e\u6210true\u7684\u8bdd":23,"\u8bbe\u7f6e\u8f93\u5165\u7c7b\u578b":23,"\u8bc4\u4f30\u4ea7\u54c1\u7684\u8d28\u91cf":13,"\u8bcd\u6027\u6807\u6ce8":12,"\u8be5\u5c42\u795e\u7ecf\u5143\u4e2a\u6570":13,"\u8be5\u6570\u636e":23,"\u8be5\u6784\u5efa\u811a\u672c\u5145\u5206\u8003\u8651\u4e86\u7f51\u7edc\u4e0d\u7a33\u5b9a\u7684\u60c5\u51b5":3,"\u8be5\u6a21\u578b\u4f9d\u7136\u662f\u4f7f\u7528\u903b\u8f91\u56de\u5f52\u5206\u7c7b\u7f51\u7edc\u7684\u6846\u67b6":13,"\u8be5\u76ee\u5f55\u4e0b\u6709\u4e24\u4e2a\u6587\u4ef6":3,"\u8be5\u811a\u672c\u7684\u4f7f\u7528\u65b9\u6cd5\u662f":3,"\u8be5image\u57fa\u4e8eubuntu":3,"\u8be5image\u7684\u6784\u5efa\u5728dock":3,"\u8be6\u60c5\u8bf7\u53c2\u8003":26,"\u8be6\u7ec6\u7684\u53c2\u6570\u89e3\u91ca\u5982\u4e0b\u9762\u8868\u683c":13,"\u8be6\u7ec6\u7684\u547d\u4ee4\u884c\u53c2\u6570\u8bf7\u53c2\u8003":26,"\u8be6\u7ec6\u7684cmake\u4f7f\u7528\u65b9\u6cd5\u53ef\u4ee5\u53c2\u8003":4,"\u8be6\u7ec6\u89c1":0,"\u8bed\u4e49\u5b8c\u5168\u76f8\u540c":1,"\u8bf4\u660e":4,"\u8bf4\u660e\u547d\u4ee4\u884c\u53c2\u6570":10,"\u8bf7\u53c2\u8003":[9,23],"\u8bf7\u53c2\u8003\u4e0b\u8282refer":23,"\u8bf7\u53c2\u8003\u4e0b\u8ff0\u6587\u7ae0":22,"\u8bf7\u5b89\u88c5cuda":10,"\u8bfb\u5165\u89e3\u6790\u8bad\u7ec3\u914d\u7f6e":26,"\u8bfb\u53d6\u6570\u636e":23,"\u8c03\u7528":4,"\u8c03\u7528\u4e00\u6b21":23,"\u8c03\u7528\u7b2c\u4e8c\u6b21\u7684\u65f6\u5019":23,"\u8d1f\u6837\u672c":13,"\u8d1f\u8d23\u591a\u673a\u8bad\u7ec3\u4e2d\u7684\u53c2\u6570\u805a\u5408\u5de5\u4f5c":16,"\u8d1f\u9762\u60c5\u7eea\u4e24\u7c7b":23,"\u8d77":1,"\u8def\u5f84\u53d8\u91cf\u4e3a":4,"\u8f83":1,"\u8f93\u5165":0,"\u8f93\u5165\u548c\u8f93\u51fa\u90fd\u662f\u5355\u5c42\u5e8f\u5217":2,"\u8f93\u5165\u548c\u8f93\u51fa\u90fd\u662f\u53cc\u5c42\u5e8f\u5217":2,"\u8f93\u5165n\u4e2a\u5355\u8bcd":13,"\u8f93\u51fa":0,"\u8f93\u51fa\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":2,"\u8f93\u51fa\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":2,"\u8f93\u51fa\u4e3an\u4e2aword":13,"\u8f93\u51fa\u5e8f\u5217\u7684\u7c7b\u578b":0,"\u8f93\u51fa\u5e8f\u5217\u7684\u8bcd\u8bed\u6570\u548c\u8f93\u5165\u5e8f\u5217\u4e00\u81f4":2,"\u8fc7\u4e86\u4e00\u4e2a\u5f88\u7b80\u5355\u7684recurr":1,"\u8fd0\u884c":[9,10],"\u8fd0\u884c\u4f7f\u7528\u7684cudnn\u5c3d\u91cf\u662f\u540c\u4e00\u4e2a\u7248\u672c":4,"\u8fd0\u884c\u8fd9\u4e2acontain":9,"\u8fd0\u884cpaddlepaddle\u7684gpu\u7248\u672c\u4e00\u5b9a\u8981\u5728\u5177\u6709cuda\u7684\u673a\u5668\u4e0a\u8fd0\u884c":3,"\u8fd1":1,"\u8fd4\u56de0":23,"\u8fd4\u56de\u4e00\u4e2alist\u6216\u8005tupl":23,"\u8fd4\u56de\u6570\u636e\u5728paddlepaddle\u4e2d\u662f\u4ec5\u4ec5\u8fd4\u56de\u4e00\u6761\u5b8c\u6574\u7684\u8bad\u7ec3\u6837\u672c":23,"\u8fd4\u56de\u7684\u987a\u5e8f\u9700\u8981\u548c":23,"\u8fd4\u56debatch_size\u7684\u5927\u5c0f":23,"\u8fd8\u4f1a":1,"\u8fd8\u662f":1,"\u8fd8\u6709":1,"\u8fd9":[1,13],"\u8fd93\u4e2a\u5b50\u6b65\u9aa4\u53ef\u914d\u7f6e\u4e3a":13,"\u8fd9\u4e00\u8fc7\u7a0b\u5bf9\u7528\u6237\u662f\u5b8c\u5168\u900f\u660e\u7684":2,"\u8fd9\u4e09\u4e2alayer\u5c06\u5b83\u5148\u53d8\u6210\u4e00\u4e2a\u5143\u7d20":1,"\u8fd9\u4e24\u5c42\u4ec5\u662f\u4e3a\u4e86\u5c55\u793a\u5b83\u4eec\u7684\u7528\u6cd5":1,"\u8fd9\u4e2a":1,"\u8fd9\u4e2a\u4e5f\u662fpaddlepaddle\u6240\u80fd\u591f\u4fdd\u8bc1\u7684shuffle\u7c92\u5ea6":23,"\u8fd9\u4e2a\u51fd\u6570\u4ee5\u4e00\u6761\u6570\u636e\u4e3a\u53c2\u6570":23,"\u8fd9\u4e2a\u51fd\u6570\u4f1a\u5728":23,"\u8fd9\u4e2a\u51fd\u6570\u5728\u521d\u59cb\u5316\u7684\u65f6\u5019\u4f1a\u88ab\u8c03\u7528":23,"\u8fd9\u4e2a\u51fd\u6570\u7684\u53c2\u6570\u662f":23,"\u8fd9\u4e2a\u521d\u59cb\u5316\u51fd\u6570\u5177\u6709\u5982\u4e0b\u53c2\u6570":23,"\u8fd9\u4e2a\u53c2\u6570\u5728\u8fd9\u4e2a\u6837\u4f8b\u91cc\u6ca1\u6709\u4f7f\u7528":23,"\u8fd9\u4e2a\u53c2\u6570\u88abpaddlepaddle\u8fdb\u7a0b\u4f20\u5165":23,"\u8fd9\u4e2a\u548c\u5728":23,"\u8fd9\u4e2a\u58f0\u660e\u57fa\u672c\u4e0a\u548cmnist\u7684\u6837\u4f8b\u4e00\u81f4":23,"\u8fd9\u4e2a\u5916\u5c42memori":1,"\u8fd9\u4e2a\u5b57\u5178\u53ef\u4ee5\u5728":23,"\u8fd9\u4e2a\u5bf9\u5e94\u5173\u7cfb\u53ef\u80fd\u4e0d\u6b63\u786e":23,"\u8fd9\u4e2a\u5bf9\u8c61\u548cprocess\u7684\u7b2c\u4e00\u4e2a\u53c2\u6570\u4e00\u81f4":23,"\u8fd9\u4e2a\u5de5\u5177\u7c7b\u63a5\u6536\u548cpydataprovider2\u4e00\u6837\u7684\u8f93\u5165\u6570\u636e":26,"\u8fd9\u4e2a\u5e8f\u5217\u6a21\u578b\u6bd4\u8f83\u590d\u6742":23,"\u8fd9\u4e2a\u5e8f\u5217\u7684\u6bcf\u4e2a\u5143\u7d20\u53c8\u662f\u4e00\u4e2a\u5e8f\u5217":2,"\u8fd9\u4e2a\u63a5\u53e3\u5e76\u4e0d\u7528\u6237\u53cb\u597d":26,"\u8fd9\u4e2a\u663e\u793a\u5668\u5f88\u68d2":13,"\u8fd9\u4e2a\u672c\u8eab\u4e0d\u662f\u4e00\u4e2a\u5f88":23,"\u8fd9\u4e2a\u6a21\u5757\u4e2d\u7684":23,"\u8fd9\u4e2a\u8bbe\u7f6e\u4e3a":23,"\u8fd9\u4e2a\u8f6f\u4ef6\u5305\u6587\u6863\u76f8\u5bf9\u5b8c\u5584":26,"\u8fd9\u4e2a\u8fc7\u7a0b\u5bf9\u7528\u6237\u4e5f\u662f\u900f\u660e\u7684":2,"\u8fd9\u4e2a\u95ee\u9898\u662fpydataprovider\u8bfb\u6570\u636e\u65f6\u5019\u7684\u903b\u8f91\u95ee\u9898":23,"\u8fd9\u4e2alayer\u7684\u8f93\u51fa\u4f1a\u4f5c\u4e3a\u6574\u4e2a":2,"\u8fd9\u4e9b\u53c2\u6570\u5305\u62ecpaddle\u5b9a\u4e49\u7684\u53c2\u6570":23,"\u8fd9\u4e9b\u53d8":4,"\u8fd9\u4e9b\u53d8\u91cf\u53ea\u5728\u7b2c\u4e00\u6b21cmake\u7684\u65f6\u5019\u6709\u6548":4,"\u8fd9\u4e9b\u53d8\u91cf\u5747\u53ef\u4ee5\u4f7f\u7528":4,"\u8fd9\u4e9b\u5b50\u53e5\u7684\u957f\u5ea6\u5f80\u5f80\u4e0d\u7b49\u957f":1,"\u8fd9\u4e9b\u6d41\u7a0b\u4e2d\u7684\u6570\u636e\u4e0b\u8f7d":13,"\u8fd9\u662f\u4e00\u79cd\u975e\u5e38\u7075\u6d3b\u7684\u6570\u636e\u7ec4\u7ec7\u65b9\u5f0f":0,"\u8fd9\u6837\u505a\u53ef\u4ee5\u907f\u514d\u5f88\u591a\u6b7b\u9501\u95ee\u9898":23,"\u8fd9\u79cd\u7c7b\u578b\u7684\u8f93\u5165\u5fc5\u987b\u901a\u8fc7":2,"\u8fd9\u884c\u7684\u4f5c\u7528\u662f\u8bbe\u7f6edataprovider\u7684\u4e00\u4e9b\u5c5e\u6027":23,"\u8fd9\u91cc":23,"\u8fd9\u91cc\u4e3e\u4f8b\u7684\u6570\u636e\u662f\u82f1\u6587\u60c5\u611f\u5206\u7c7b\u7684\u6570\u636e":23,"\u8fd9\u91cc\u4ee5":13,"\u8fd9\u91cc\u4ee5mnist\u624b\u5199\u8bc6\u522b\u4e3a\u4f8b":23,"\u8fd9\u91cc\u53ef\u4ee5\u53c2\u8003paddle\u7684":21,"\u8fd9\u91cc\u6211\u4eec\u4f7f\u7528\u7b80\u5355\u7684\u6587\u672c\u6587\u4ef6\u8868\u793amnist\u56fe\u7247":23,"\u8fd9\u91cc\u6307\u5b9a\u8bcd\u5178":13,"\u8fd9\u91cc\u6ca1\u6709\u4ecb\u7ecd\u591a\u673a\u5206\u5e03\u5f0f\u8bad\u7ec3":13,"\u8fd9\u91cc\u7684":23,"\u8fd9\u91cc\u7684\u8f93\u5165\u7279\u5f81\u662f\u8bcdid\u7684\u5e8f\u5217":23,"\u8fd9\u91cc\u8981\u6ce8\u610f\u9884\u6d4b\u6570\u636e\u901a\u5e38":26,"\u8fd9\u91cc\u8bbe\u7f6e\u7684\u662f\u8fd4\u56de\u4e00\u4e2a":23,"\u8fd9\u91cc\u8bf4\u660e\u4e86\u8bad\u7ec3\u6570\u636e\u662f":23,"\u8fd9\u91cc\u91c7\u7528adam\u4f18\u5316\u65b9\u6cd5":13,"\u8fdb\u5165\u8be5\u6e90\u7801\u76ee\u5f55":3,"\u8fdb\u5165docker":9,"\u8fdc\u7a0b\u8bbf\u95ee":9,"\u8fde\u63a5":2,"\u8fde\u63a5\u8bf7\u53c2\u8003":13,"\u9002\u4e2d":1,"\u9009":1,"\u9009\u62e9":1,"\u9009\u62e9\u666e\u901acpu\u7248\u672c\u7684devel\u7248\u672c\u7684imag":9,"\u9009\u9879":4,"\u901a\u5e38\u6839\u636e\u4efb\u52a1\u9700\u6c42\u8fdb\u884c\u4e0d\u540c\u8bbe\u7f6e":1,"\u901a\u77e5":1,"\u901a\u8fc7\u4e24\u4e2a\u5d4c\u5957\u7684":2,"\u901a\u8fc7\u591a\u7ec4\u8bed\u4e49\u76f8\u540c\u7684\u5355\u53cc\u5c42rnn\u914d\u7f6e":1,"\u901a\u8fc7\u5f15\u7528memory\u5f97\u5230\u8fd9\u4e2alayer\u4e0a\u4e00\u4e2a\u65f6\u523b\u7684\u8f93\u51fa":2,"\u901a\u8fc7\u5f15\u7528memory\u5f97\u5230\u8fd9\u4e2alayer\u4e0a\u4e00\u4e2a\u65f6\u523b\u8f93\u51fa":2,"\u901a\u8fc7\u7f16\u8bd1\u65f6\u6307\u5b9a\u8def\u5f84\u6765\u5b9e\u73b0\u5f15\u7528\u5404\u79cdbla":4,"\u901a\u8fc7data":2,"\u903b\u8f91\u56de\u5f52":13,"\u9053\u6b49":1,"\u9069":1,"\u90a3\u4e48":[2,23],"\u90a3\u4e480\u5c42\u5e8f\u5217\u5373\u4e3a\u4e00\u4e2a\u8bcd\u8bed":2,"\u90a3\u4e48\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d":22,"\u90a3\u4e48\u5bf9\u5e94\u7684dataprovider\u65e2\u4e3a":23,"\u90a3\u4e48\u8fd9\u4e2a\u4e0b\u8f7d\u7ebf\u7a0b\u5c06\u4f1a\u5173\u95ed":3,"\u90a3\u4e48paddlepaddle\u4f1a\u7c97\u7565\u7684\u6839\u636elayer\u7684\u58f0\u660e\u987a\u5e8f":23,"\u90fd":1,"\u90fd\u4f20\u9012\u7ed9process\u51fd\u6570":23,"\u90fd\u662f\u5bf9layer1\u5143\u7d20\u7684\u62f7\u8d1d":0,"\u914d\u7f6e":1,"\u914d\u7f6e\u4e86":23,"\u914d\u7f6e\u53c2\u6570\u914d\u7f6e\u7ed9dataprovider\u7684":23,"\u914d\u7f6e\u6587\u4ef6":13,"\u914d\u7f6eapi":0,"\u9152\u5e97":1,"\u91cc\u4f1a\u7ee7\u7eed\u5b89\u88c5":10,"\u91cc\u63d0\u4f9b\u4e86\u6570\u636e\u4e0b\u8f7d\u811a\u672c":13,"\u91cc\u9762\u8bfb\u53d6":23,"\u91cf\u4e5f\u53ef\u4ee5\u901a\u8fc7\u8c03\u7528cmake\u547d\u4ee4\u524d\u901a\u8fc7\u73af\u5883\u53d8\u91cf\u6307\u5b9a":4,"\u9488\u5bf9\u672c\u95ee\u9898":13,"\u94fe\u63a5\u4f55\u79cdblas\u7b49\u7b49":4,"\u9519\u8bef\u7387":13,"\u95f4\u63a5\u4f7f\u7528":1,"\u95f4\u9694":23,"\u9664\u4e86":23,"\u9664\u4e86boot":1,"\u9664\u8fc7data\u5c42":13,"\u9700\u8981\u53c2\u8003":9,"\u9700\u8981\u652f\u6301avx\u6307\u4ee4\u96c6\u7684cpu":9,"\u9700\u8981\u6ce8\u610f":23,"\u9700\u8981\u6ce8\u610f\u7684\u662f":[4,10],"\u9700\u8981\u9075\u5faa\u4ee5\u4e0b\u7ea6\u5b9a":2,"\u9884\u6d4b\u6570\u636e\u6307\u5b9atest":13,"\u9884\u6d4b\u7ed3\u679c\u4ee5\u6587\u672c\u7684\u5f62\u5f0f\u4fdd\u5b58\u5728":13,"\u9884\u6d4b\u811a\u672c":13,"\u9884\u6d4bid":13,"\u989d\u5916\u7684\u53c2\u6570":13,"\u9996\u5148":1,"\u9996\u5148\u5217\u4e3e\u903b\u8f91\u56de\u5f52\u7f51\u7edc":13,"\u9996\u5148\u6211\u4eec\u5c06\u8fd9\u4e2a\u6570\u636e\u6587\u4ef6":23,"\u9996\u5148\u8bf7\u53c2\u8003":13,"\u9aa43":13,"\u9ed8\u8ba4\u4e00\u4e2apass\u4fdd\u5b58\u4e00\u6b21\u6a21\u578b":13,"\u9ed8\u8ba4\u4e0d\u8bbe\u7f6e":2,"\u9ed8\u8ba4\u4e3a\u7b2c\u4e00\u4e2a\u8f93\u5165":2,"\u9ed8\u8ba4\u503c":[0,4],"\u9ed8\u8ba4\u521d\u59cb\u72b6\u4e3a0":2,"\u9ed8\u8ba4\u5355\u4f4d\u662fbyte":3,"\u9ed8\u8ba4\u60c5\u51b5\u4e0b\u4e00\u6761\u6570\u636e":23,"adamax\u7b49":13,"amazon\u7535\u5b50\u4ea7\u54c1\u8bc4\u8bba\u6570\u636e":13,"api\u9884\u6d4b":13,"argument\u4f20\u5165":23,"argument\u5f62\u5f0f\u4f20\u5165":23,"atlas\u5e93\u7684\u8def\u5f84":4,"batches\u8bbe\u7f6e\u6bcf\u9694\u591a\u5c11batch\u4fdd\u5b58\u4e00\u6b21\u6a21\u578b":13,"bool\u53c2\u6570":23,"case":[13,25],"cd\u5230\u542b\u6709dockerfile\u7684\u8def\u5f84\u4e2d":9,"check\u662ffalse\u7684\u8bdd":23,"cmake\u53ef\u4ee5\u5c06cmake\u9879\u76ee\u6587\u4ef6":4,"cmake\u662f\u4e00\u4e2a\u8de8\u5e73\u53f0\u7684\u7f16\u8bd1\u811a\u672c":4,"cmake\u7684\u5b98\u65b9\u6587\u6863":4,"cmake\u7f16\u8bd1\u65f6\u4f1a\u9996\u5148\u5728\u7cfb\u7edf\u8def\u5f84":4,"container\u540e":9,"cpu\u7248\u672c":9,"cuda\u76f8\u5173\u7684driver\u548c\u8bbe\u5907\u6620\u5c04\u8fdbcontainer\u4e2d":9,"d\u547d\u4ee4\u5373\u53ef":4,"d\u547d\u4ee4\u6307\u5b9a":4,"dataprovider\u521b\u5efa\u7684\u65f6\u5019\u6267\u884c":23,"dataprovider\u53ef\u4ee5\u662f":23,"dataprovider\u63d0\u4f9b\u4e86\u4e24\u79cd\u7b80\u5355\u7684cache\u7b56\u7565":23,"dataprovider\u7684\u5177\u4f53\u7528\u6cd5\u548c\u5982\u4f55\u5b9e\u73b0\u4e00\u4e2a\u65b0\u7684dataprovid":22,"decoder\u5faa\u73af\u5c55\u5f00\u7684\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u4f1a\u5f15\u7528\u5168\u90e8\u7ed3\u679c":2,"decoder\u63a5\u53d7\u4e24\u4e2a\u8f93\u5165":2,"decoder\u6bcf\u6b21\u9884\u6d4b\u4ea7\u751f\u4e0b\u4e00\u4e2a\u6700\u53ef\u80fd\u7684\u8bcd\u8bed":2,"decoer\u67b6\u6784":2,"devel\u548cdemo":9,"dim\u7684\u65b0\u7684\u5411\u91cf":13,"dim\u7ef4\u5ea6\u5411\u91cf":13,"docker\u662f\u4e00\u4e2a\u57fa\u4e8e\u5bb9\u5668\u7684\u8f7b\u91cf\u7ea7\u865a\u62df\u73af\u5883":9,"docker\u7684\u5b98\u65b9\u6587\u6863":9,"dockerfile\u548cbuild":3,"dockerfile\u662fdock":3,"dockerfile\u7684\u6587\u6863":9,"dockerfile\u7684\u6700\u4f73\u5b9e\u8df5":9,"driver\u6dfb\u52a0\u5230ld_library_path\u4e2d":10,"elec\u6d4b\u8bd5\u96c6":13,"embedding\u6a21\u578b\u9700\u8981\u7a0d\u5fae\u6539\u53d8\u6570\u636e\u63d0\u4f9b\u7684\u811a\u672c":13,"encoder\u548cdecoder\u53ef\u4ee5\u662f\u80fd\u591f\u5904\u7406\u5e8f\u5217\u7684\u4efb\u610f\u795e\u7ecf\u7f51\u7edc\u5355\u5143":2,"encoder\u8f93\u51fa":2,"export":[4,9,10],"f\u4ee3\u8868\u4e00\u4e2a\u6d6e\u70b9\u6570":23,"float":23,"generator\u4fbf\u4f1a\u5b58\u4e0b\u5f53\u524d\u7684\u4e0a\u4e0b\u6587":23,"generator\u7684\u4e0a\u4e0b\u6587\u4e2d\u5c3d\u91cf\u7559":23,"generator\u81f3\u5c11\u8c03\u7528\u4e24\u6b21\u624d\u4f1a\u77e5\u9053\u662f\u5426\u505c\u6b62":23,"gpu\u53cc\u7f13\u5b58":23,"gpu\u7248\u672c":9,"gpu\u7248\u672c\u4e8c\u8fdb\u5236":4,"group\u548c\u5355\u5c42\u5e8f\u5217\u7684\u51e0\u4e4e\u4e00\u6837":1,"group\u5916":1,"gru\u6a21\u578b":13,"gru\u6a21\u578b\u914d\u7f6e":13,"i\u4ee3\u8868\u4e00\u4e2a\u6574\u6570":23,"id\u4e3a0\u7684\u6982\u7387":13,"id\u4e3a1\u7684\u6982\u7387":13,"image\u6784\u5efa\u6e90\u7801\u653e\u7f6e\u5728":3,"image\u7684\u4e3b\u8981\u63cf\u8ff0\u6587\u4ef6":3,"image\u7684\u4e3b\u8981\u6784\u5efa\u6b65\u9aa4":3,"image\u7684\u6784\u5efa\u6b65\u9aa4":3,"import":[13,23,26],"include\u4e0b\u9700\u8981\u5305\u542bcbla":4,"include\u4e0b\u9700\u8981\u5305\u542bmkl":4,"init_hook\u53ef\u4ee5\u4f20\u5165\u4e00\u4e2a\u51fd\u6570":23,"int":[1,13,23],"key\u662fdata_layer\u7684\u540d\u5b57":23,"label\u662finteg":1,"layer1\u5fc5\u987b\u662f\u4e00\u4e2a0\u5c42\u5e8f\u5217":0,"layer1\u5fc5\u987b\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"layer2\u4e00\u81f4":0,"layer2\u53ef\u4ee5\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":0,"layer2\u5fc5\u987b\u662f\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":0,"layer\u4e0d\u5173\u5fc3\u6570\u636e\u662f\u5426\u662f\u5e8f\u5217\u683c\u5f0f":1,"layer\u4e0d\u80fd\u94fe\u63a5\u5916\u5c42\u7684\u8fd9\u4e2amemori":1,"layer\u4f20\u7ed9\u4e0b\u4e00\u4e2a\u5b50\u53e5\u7684memori":1,"layer\u4f5c\u4e3a\u8f93\u51fa":26,"layer\u540e":1,"layer\u548caverag":1,"layer\u548cembed":1,"layer\u548clstmemori":1,"layer\u5c42":1,"layer\u62ff\u5230\u7684\u7528\u6237\u8f93\u5165":2,"layer\u6587\u6863":13,"layer\u7684\u4f7f\u7528\u793a\u4f8b\u5982\u4e0b":0,"ld_library_path\u7b49\u7b49":10,"ld_library_path\u91cc\u9762\u627e\u4e0d\u5230\u8fd9\u4e9b\u52a8\u6001":10,"lib\u4e0b\u9700\u8981\u5305\u542bcblas\u548catlas\u4e24\u4e2a\u5e93":4,"lib\u4e0b\u9700\u8981\u5305\u542bcblas\u5e93":4,"lib\u4e0b\u9700\u8981\u5305\u542bopenblas\u5e93":4,"lib\u76ee\u5f55\u4e0b\u9700\u8981\u5305\u542b":4,"list\u4e0d\u8bbe\u7f6e":22,"list\u4e2d":[22,23],"list\u4e2d\u7684\u4e00\u884c":23,"list\u4e2d\u7684\u6bcf\u4e00\u884c":23,"list\u4e3a\u7eaf\u6587\u672c\u6587\u4ef6":22,"list\u4e5f\u53ef\u4ee5\u653e\u7f6ehdfs\u6587\u4ef6\u8def\u5f84":22,"list\u5199\u5165\u90a3\u4e2a\u6587\u672c\u6587\u4ef6\u7684\u5730\u5740":23,"list\u5373\u4e3a":23,"list\u548ctest":22,"list\u5747\u4e3a\u672c\u5730\u7684\u4e24\u4e2a\u6587\u4ef6":22,"list\u6307\u5b9a\u7684\u6570\u636e":13,"list\u7684\u4f4d\u7f6e":13,"list\u82e5\u5e72\u6570\u636e\u6587\u4ef6\u8def\u5f84\u7684\u67d0\u4e00\u4e2a\u8def\u5f84":23,"lstm\u6a21\u578b\u7b49":13,"lstm\u6a21\u578b\u914d\u7f6e":13,"make\u548cmak":5,"mem\u4f5c\u4e3a\u5185\u5c42memory\u7684\u521d\u59cb\u72b6\u6001":1,"mem\u662f\u4e00\u4e2a\u5b50\u53e5\u7684\u6700\u540e\u4e00\u4e2a\u5411\u91cf":1,"memory\u4e0d\u80fd\u72ec\u7acb\u5b58\u5728":2,"memory\u53ea\u80fd\u5728":2,"memory\u6307\u5411\u4e00\u4e2alay":2,"memory\u7684\u521d\u59cb\u72b6\u6001":2,"memory\u7684\u66f4\u591a\u8ba8\u8bba\u8bf7\u53c2\u8003\u8bba\u6587":2,"memory\u7684i":2,"memory\u9ed8\u8ba4\u521d\u59cb\u5316\u4e3a0":2,"mkl\u7684\u8def\u5f84":4,"mkl_sequential\u548cmkl_intel_lp64\u4e09\u4e2a\u5e93":4,"mnist\u662f\u4e00\u4e2a\u5305\u542b\u6709":23,"movielens\u6570\u636e\u96c6":12,"movielens\u8bc4\u5206\u56de\u5f52":12,"name\u90fd\u662f":9,"osx\u6216\u8005\u662fwindows\u673a\u5668":9,"osx\u7684\u5b89\u88c5\u6587\u6863":9,"paddle\u5b9a\u4e49\u7684\u53c2\u6570\u5305\u62ec":23,"paddle\u7684":10,"paddlepaddle\u4e2d":[0,2],"paddlepaddle\u4f7f\u7528\u8fd0\u884c\u65f6\u52a8\u6001\u8fde\u63a5cuda\u7684so":10,"paddlepaddle\u4fdd\u7559\u6dfb\u52a0\u53c2\u6570\u7684\u6743\u529b":23,"paddlepaddle\u53ef\u4ee5\u4f7f\u7528":4,"paddlepaddle\u53ef\u4ee5\u8bfb\u53d6python\u5199\u7684\u4f20\u8f93\u6570\u636e\u811a\u672c":13,"paddlepaddle\u5728\u8fd0\u884c\u65f6\u627e\u4e0d\u5230\u5bf9\u5e94\u7684config\u6587\u4ef6":10,"paddlepaddle\u5c06train":23,"paddlepaddle\u63a8\u8350\u4f7f\u7528docker\u8fdb\u884cpaddlepaddle\u7684\u90e8\u7f72\u548c":9,"paddlepaddle\u63d0\u4f9b\u4e86docker\u7684\u4f7f\u7528\u955c\u50cf":9,"paddlepaddle\u63d0\u4f9b\u6570\u4e2a\u9884\u7f16\u8bd1\u7684\u4e8c\u8fdb\u5236\u6765\u8fdb\u884c\u5b89\u88c5":8,"paddlepaddle\u63d0\u4f9b\u7684\u955c\u50cf\u5e76\u4e0d\u5305\u542b\u4efb\u4f55\u547d\u4ee4\u8fd0\u884c":9,"paddlepaddle\u7684\u6570\u636e\u5305\u62ec\u56db\u79cd\u4e3b\u8981\u7c7b\u578b":23,"paddlepaddle\u7684\u7f16\u8bd1\u9009\u9879\u53ef\u4ee5\u5728\u8c03\u7528cmake\u7684\u65f6\u5019\u8bbe\u7f6e":4,"paddlepaddle\u7684\u7f16\u8bd1\u9009\u9879\u662f\u53ef\u4ee5\u63a7\u5236paddlepaddle\u751f\u6210cpu":4,"paddlepaddle\u7684dock":3,"paddlepaddle\u7684python\u9884\u6d4b\u63a5\u53e3":25,"paddlepaddle\u7684ubuntu\u5b89\u88c5\u5305\u5206\u4e3a\u56db\u4e2a\u7248\u672c":10,"paddlepaddle\u76ee\u524d\u4f7f\u7528swig\u5bf9\u5176\u5e38\u7528\u7684\u9884\u6d4b\u63a5\u53e3\u8fdb\u884c\u4e86\u5c01\u88c5":26,"paddlepaddle\u76ee\u524d\u652f\u6301\u4f7f\u7528deb\u5305\u5b89\u88c5":10,"paddlepaddle\u8d1f\u8d23\u5b8c\u6210\u4fe1\u606f\u548c\u68af\u5ea6\u5728\u65f6\u95f4\u5e8f\u5217\u4e0a\u7684\u4f20\u64ad":2,"paddlepaddle\u8d1f\u8d23\u5b8c\u6210\u4fe1\u606f\u548c\u8bef\u5dee\u5728\u65f6\u95f4\u5e8f\u5217\u4e0a\u7684\u4f20\u64ad":2,"paddlepaddle\u8fd0\u884c\u65f6\u5982\u679c\u6ca1\u6709\u5bfb\u627e\u5230cuda\u7684driv":10,"paddlepaddle\u9700\u8981\u7528\u6237\u5728\u7f51\u7edc\u914d\u7f6e":22,"period\u8bbe\u7f6e\u6253\u5370\u53c2\u6570\u4fe1\u606f\u7b49":13,"process\u51fd\u6570":23,"process\u51fd\u6570\u662f\u5b9e\u73b0\u6570\u636e\u8f93\u5165\u7684\u4e3b\u51fd\u6570":23,"process\u51fd\u6570\u8c03\u7528\u591a\u6b21":23,"pserver\u4e3apaddlepaddle\u7684paramet":16,"pserver\u7684\u547d\u4ee4\u884c\u53c2\u6570":16,"pserver\u7ec4\u5408\u4f7f\u7528":16,"py\u6587\u4ef6\u7ed9\u51fa\u4e86\u5b8c\u6574\u4f8b\u5b50":13,"pydataprovider2\u4f1a\u5c3d\u91cf\u4f7f\u7528\u5185\u5b58":23,"pydataprovider2\u6587\u6863":26,"pydataprovider2\u7684\u4f7f\u7528":22,"pydataprovider\u662fpaddlepaddle\u4f7f\u7528python\u63d0\u4f9b\u6570\u636e\u7684\u63a8\u8350\u63a5\u53e3":23,"python\u5305":9,"python\u53ef\u4ee5\u89e3\u9664\u6389\u5185\u90e8\u53d8\u91cf\u7684\u5f15\u7528":23,"python\u7684":9,"python\u7684swig\u63a5\u53e3\u53ef\u4ee5\u65b9\u4fbf\u8fdb\u884c\u9884\u6d4b\u548c\u5b9a\u5236\u5316\u8bad\u7ec3":4,"return":[1,13,23],"rnn\u603b\u662f\u5f15\u7528\u4e0a\u4e00\u65f6\u523b\u9884\u6d4b\u51fa\u7684\u8bcd\u7684\u8bcd\u5411\u91cf":2,"search\u7684\u751f\u6210":1,"seq\u53c2\u6570\u5fc5\u987b\u4e3afals":2,"seq\u540e":1,"seq\u5c42":1,"seq\u7684\u4f7f\u7528\u793a\u4f8b\u5982\u4e0b":0,"seq\u7c7b\u4f3c":0,"sequence\u7c7b\u578b":1,"server\u8fdb\u7a0b":16,"sh\u662fdocker":3,"shuffle\u8bad\u7ec3\u6570\u636e":23,"slot\u662finteg":1,"softmax\u8f93\u51fa":13,"state\u505a\u4e86\u4e00\u4e2a\u5168\u94fe\u63a5":1,"step\u4e2d":1,"step\u51fd\u6570\u4e2d\u7684memori":2,"step\u51fd\u6570\u5185\u90e8\u53ef\u4ee5\u81ea\u7531\u7ec4\u5408paddlepaddle\u652f\u6301\u7684\u5404\u79cdlay":2,"step\u7684recurr":1,"string\u7684\u683c\u5f0f\u6253\u5370\u51fa\u6765":16,"subseq\u7684\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2a0\u5c42\u5e8f\u5217":0,"swig_paddle\u63a5\u53d7\u7684\u539f\u59cb\u6570\u636e\u662fc":26,"tag\u5206\u522b\u4e3a":9,"train\u5373\u4e3apaddlepaddle\u7684\u8bad\u7ec3\u8fdb\u7a0b":16,"train\u5b8c\u6210\u5355\u673a\u591a\u663e\u5361\u591a\u7ebf\u7a0b\u7684\u8bad":16,"train\u7684\u547d\u4ee4\u884c\u53c2\u6570":16,"true":1,"true\u7684memory\u65f6":1,"types\u7684\u8be6\u7ec6\u7528\u6cd5":1,"ubuntu\u7684deb\u5b89\u88c5\u5305\u7b49":8,"v2\u4e4b\u540e\u7684\u4efb\u4f55\u4e00\u4e2acudnn\u7248\u672c\u6765\u7f16\u8bd1\u8fd0\u884c":4,"value\u5373\u4e3asoftmax\u5c42\u7684\u8f93\u51fa":26,"value\u662f\u7279\u5f81\u503c":23,"value\u7c7b\u578b":1,"var":9,"vector\u8868\u793a\u7684\u6bcf\u4e2a\u5355\u8bcd":13,"version\u53ef\u4ee5\u6253\u5370\u51fapaddle\u7684\u7248\u672c\u4fe1\u606f\u548c\u7f16\u8bd1\u7684\u9009\u9879":21,"version\u53ef\u4ee5\u6253\u5370\u51fapaddlepaddle\u7684\u7248\u672c\u548c\u7f16\u8bd1\u65f6\u4fe1\u606f":16,"version\u7684\u547d\u4ee4\u884c\u53c2\u6570":16,"yield\u6587\u672c\u4fe1\u606f\u548c\u7c7b\u522bid":13,__main__:26,__name__:26,abov:23,act:[1,13],act_typ:13,activ:13,adadelta:13,adagrad:13,adam:13,adamoptim:13,afi:23,agg_level:[0,1],aggregatelevel:[0,1],all:[2,23],allow:13,alreadi:10,also:13,append:[1,23],apt:[9,10],arg:[3,13,23],around:23,arrai:26,assert:26,atla:4,atlas_root:4,averag:1,avg:13,avgcost:13,avgpool:[0,1,13],avx:9,bag:13,baidu:[9,10],batch:13,batch_siz:[1,13],batchsiz:1,beam:1,beam_search:2,bias_attr:1,binari:13,bla:4,bool:13,boot:2,boot_lay:1,both:13,bow:13,build:[3,9],cach:[13,22],cache_pass_in_mem:[13,23],cachetyp:[13,23],calc_batch_s:23,call:13,can:13,can_over_batch_s:23,cat:9,categori:13,check:[1,10,23],check_fail_continu:23,chines:12,chpasswd:9,classif:13,classification_cost:[1,13],classification_error_evalu:13,close:23,cmake:4,cmd:9,cnn:13,code:[3,23,26],com:[9,10],comment:[1,13],compil:[10,21],conf:[1,26],config:[10,13],config_arg:13,config_pars:26,connect:13,contain:[13,23],context:23,context_len:13,context_start:13,convert:[13,23,26],couldn:10,cpp:[10,13],cpu:[9,10,23],cpuinfo:9,createfromconfigproto:26,cross:13,cuda_so:9,cudastat:10,cudasuccess:10,cudnn:4,cudnn_root:4,cudnnv5:4,current:[13,23],currentcost:13,currentev:13,dalla:23,data:[1,10],data_config:26,data_initialz:13,data_lay:[1,13,23],dataprovid:13,dataprovider_bow:13,dataprovider_emb:13,dataproviderconvert:26,dataset:13,deb:10,debian:10,decod:2,decor:23,def:[1,13,23,26],defin:[13,23],define_py_data_sources2:[13,23],delar:13,demo:[9,13],dense_vector:[23,26],describ:13,descript:25,detail:25,dev:9,devel:9,devic:9,devices:9,dict:[13,23],dict_dim:1,dict_fil:[1,13],dictionai:13,dictionari:[13,23],dictrionari:13,differ:13,dim:13,dimens:13,dir:13,doc:26,documentari:23,dpkg:10,driver:10,dso_handl:10,dtype:26,dump_config:16,dure:[13,23],dynam:23,each:[13,23],each_pixel_str:23,each_sequence:[0,1],each_word:23,echo:9,either:13,els:[1,9,13],emb:[1,13],emb_group:1,embed:12,embedding_lay:[1,13],entropi:13,enumer:13,equal:1,error:[10,13],error_clipping_threshold:1,etc:9,eval:13,exampl:13,expand:[0,1],expand_a:[0,1],expand_level:[0,1],expandlevel:[0,1],expose:9,extralayerattribut:1,f0831:10,fail:[1,10],fals:13,fc_layer:[1,13],fdata:1,featur:[13,23],festiv:23,file:[13,23],file_list:23,file_nam:[1,13],filenam:23,fill:13,find:10,first:[0,13],float32:26,fly:13,forwardtest:26,framework:13,from:[2,9,13,23,26],from_sequence:[0,1],from_timestep:0,full_matrix_project:1,fulli:13,func:23,gate_act:1,gdebi:10,gener:[13,23],generatedinput:2,get:[9,10,13,23],get_config_arg:13,get_data:13,github:10,give:23,given:13,globe:23,gpu:[9,10],gradient_clipping_threshold:13,gradientmachin:26,grep:9,group:1,group_input:1,gru:13,gru_siz:13,gserver:1,hassubseq:1,help:26,hidden_dim:1,hierach:2,hint:26,hl_cuda_devic:10,hl_dso_load:10,hook2:1,hook:1,host:9,hot:13,hous:23,howardjohnson:1,http:10,ignor:23,imag:9,imagenet:12,img:23,inarg:26,includ:13,init:13,init_hook:[1,13,22],init_model_path:13,initi:[13,23],initpaddl:26,inner_mem:1,inner_rnn_output:1,inner_rnn_st:1,inner_step:1,input:[0,1,2,13,23],input_typ:[1,13,22],instal:5,insuffici:10,integ:[13,23],integer_sequ:23,integer_valu:[1,13,23],integer_value_sequ:[1,13],integer_value_sub_sequ:1,invok:23,is_predict:13,is_train:23,isinst:26,iterat:23,job:13,join:1,kernel:9,kwarg:[1,13,23],l2regular:13,label:[1,13,23],label_dim:[1,13],label_list:1,lake:23,last:[0,1],later:13,latest:[3,9],layer1:0,layer2:0,layer:[0,1,2,13],ld_library_path:10,learning_method:13,learning_r:13,len:[1,13,23],level:2,lib64:[9,10],lib:4,libcuda:9,libnvidia:9,librari:10,line:1,link:2,list:[13,22,23],load_data_arg:26,loadparamet:26,local:[4,10],log_period:13,logger:23,look:[13,23],loss:13,lowest_dl_speed:3,lstm:[1,13],lstm_averag:1,lstm_expand:1,lstm_group:1,lstm_group_input:1,lstm_input:1,lstm_last:1,lstm_layer_attr:1,lstm_nest_group:1,lstm_output:1,lstm_size:13,lstmemori:1,lstmemory_group:1,mac:9,machin:2,main:26,maintainer:9,make:[10,23],make_diagram:16,maxid:13,maxid_lay:13,maxpool:0,mean:13,mem:1,memori:1,merge_model:16,method:23,min_pool_s:23,mixed_lay:1,mkdir:9,mkl:4,mkl_core:4,mkl_root:4,mnist:23,mnist_model:26,mnist_provid:23,mnist_train:23,model_config:26,modul:[13,23],momentum:13,movi:23,must:10,name:[1,9,13,23],necessari:13,need:13,neg:[13,23],nest:1,net:9,neural:2,next:23,no_cache:23,no_sequence:23,noavx:[9,10],none:[13,23,26],normal:9,note:10,now:2,nullptr:10,num:13,num_pass:13,nvidia:9,obj:[13,23],object:[13,23],off:[3,4,10,21],omit:13,on_init:23,onli:[2,13],open:[1,13,23],openbla:4,openblas_root:4,openssh:9,opt:4,other:13,out:[1,2],outer:1,outer_mem:1,outer_rnn_st:1,outer_step:1,outlin:25,output:[1,13],outsid:23,paddl:[1,3,9,10,13,16],paddle_gpu:3,paddle_ssh:9,paddle_ssh_machin:9,paddledev:9,paddlepaddl:[9,10,21,26],paramet:13,parse_config:26,pass:[13,23],path:[10,13],period:13,permitrootlogin:9,pixel:23,pixels_float:23,pixels_str:23,place:23,pleas:10,pool:0,pool_siz:23,pooling_typ:[0,1,13],posit:[13,23],pred:13,predict_output_dir:13,predict_sampl:26,preprocess:13,print:26,proc:9,process2:1,process:[1,13,23],process_pr:13,process_seq:1,process_subseq:1,properli:13,provid:1,pull:9,put:13,py_paddl:[9,26],pydataprovid:22,pydataprovider2:[13,23,26],pydataproviderwrapp:13,python:13,quick_start:13,rang:13,rank:13,rare:23,read:[13,23],real_process:23,recurrent_group:[1,2],refer:22,reference_cblas_root:4,reffer:4,regular:13,releas:10,repres:13,represent:13,resnet:12,result:[13,23],revers:2,rmsprop:13,rnn:2,rnn_data_provid:1,rnn_state:1,roce:9,root:9,run:9,runtim:[10,23],same:[13,23],sampl:[13,23],save:[13,23],save_dir:13,saw:23,sbin:9,script:3,second:13,sed:9,see:13,sentenc:23,sentiment:23,sentimental_provid:23,separ:13,seq:[0,1],seq_pool:0,seq_typ:23,seqlastin:1,sequel:23,sequenc:[1,2],sequence:23,sequence_conv_pool:13,sequence_layer_group:1,sequence_nest_layer_group:1,sequence_nest_rnn:1,sequence_nest_rnn_readonly_memori:1,sequence_rnn:1,sequencegen:1,sequencestartposit:1,sequencetyp:23,server:9,set:[1,13,23],setup:13,should:2,should_shuffl:23,sigmoidactiv:1,simple_gru:13,simple_lstm:13,size:[1,13,23],softmax:13,softmaxactiv:[1,13],sourc:13,spars:13,sparse_binary_vector:[13,23],sparse_float_vector:23,specifi:[10,13],split:[1,13,23],src_root:26,ssh:9,sshd:9,sshd_config:9,stat:13,state:2,state_act:1,staticinput:2,step:[1,2],stop:9,store:13,string:23,strip:[1,13],structur:13,stun:23,sub:1,sub_sequence:23,subseq:[0,2],subsequenceinput:1,sudo:10,support:9,sure:10,swig_paddl:26,tag:3,take:23,tanhactiv:1,tbd:[1,24],team:9,test:[1,13,22],test_data:26,test_list:[13,23],test_recurrentgradientmachin:1,text:[13,23],text_conv:13,them:13,thi:[13,23],thing:23,timestep:0,tmp:23,tour_train_wdseg:1,train:10,train_list:[13,23],trainer:[13,23,26],trainer_config:[13,22,23,26],trainer_config_help:[13,23],trainerintern:13,trainermain:10,travel:23,trn:13,tst:13,turn:2,two:13,txt:[13,23],type:[13,23],unk_idx:13,updat:9,use:[13,25],use_dynamic_ord:23,use_gpu:[13,26],usepam:9,user:13,usr:[4,9,10],valid:10,valu:[1,13,23,26],version:[9,10],via:10,want:23,what:13,when:23,which:13,whole:23,wilder:23,window:9,with_avx:[4,10,21],with_doc:4,with_doc_cn:4,with_doubl:[10,21],with_double:4,with_dso:4,with_gflag:[10,21],with_gflags:4,with_glog:[4,10,21],with_gpu:[3,4,10,21],with_metric_learn:[10,21],with_predict_sdk:[10,21],with_python:[4,10,21],with_rdma:[4,10,21],with_style_check:4,with_swig_py:4,with_testing:4,with_tim:[10,21],with_timer:4,without:9,wonder:23,word:[1,2,12],word_dict:[1,13],word_dim:[1,13],word_id:23,word_slot:1,word_slot_list:1,word_vector:13,xarg:9,yield:[1,13,23],you:[10,23],your_host_machine:9},titles:["\u652f\u6301\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165\u7684Layer","\u53cc\u5c42RNN\u914d\u7f6e\u4e0e\u793a\u4f8b","Recurrent Group\u6559\u7a0b","\u6784\u5efaPaddlePaddle Docker Image","\u8bbe\u7f6ePaddlePaddle\u7684\u7f16\u8bd1\u9009\u9879","\u4f7f\u7528cmake\u7f16\u8bd1PaddlePaddle","\u5b89\u88c5\u7f16\u8bd1PaddlePaddle\u9700\u8981\u7684\u4f9d\u8d56","make\u548cmake install","\u7f16\u8bd1\u4e0e\u5b89\u88c5","\u5b89\u88c5PaddlePaddle\u7684Docker\u955c\u50cf","\u4f7f\u7528deb\u5305\u5728Ubuntu\u4e0a\u5b89\u88c5PaddlePaddle","\u96c6\u7fa4\u8bad\u7ec3","\u4f7f\u7528\u793a\u4f8b","PaddlePaddle\u5feb\u901f\u5165\u95e8\u6559\u7a0b","PaddlePaddle\u6587\u6863","&lt;no title&gt;","\u547d\u4ee4\u884c\u53c2\u6570","&lt;no title&gt;","&lt;no title&gt;","paddle pserver\u7684\u547d\u4ee4\u884c\u53c2\u6570","paddle train\u7684\u547d\u4ee4\u884c\u53c2\u6570","paddle version\u7684\u547d\u4ee4\u884c\u53c2\u6570","PaddlePaddle\u7684\u6570\u636e\u63d0\u4f9b(DataProvider)\u4ecb\u7ecd","PyDataProvider2\u7684\u4f7f\u7528","\u81ea\u5b9a\u4e49\u4e00\u4e2aDataProvider","\u7528\u6237\u63a5\u53e3","PaddlePaddle\u7684Python\u9884\u6d4b\u63a5\u53e3"],titleterms:{"\u4e0b\u8f7d\u548c\u8fd0\u884cdocker\u955c\u50cf":9,"\u4ecb\u7ecd":22,"\u4f18\u5316\u7b97\u6cd5":13,"\u4f7f\u7528\u6307\u5357":14,"\u4f7f\u7528\u6982\u8ff0":13,"\u4f7f\u7528\u793a\u4f8b":12,"\u4f7f\u7528\u811a\u672c\u6784\u5efapaddlepaddl":3,"\u4f7f\u7528cmake\u7f16\u8bd1paddlepaddl":5,"\u4f7f\u7528deb\u5305\u5728ubuntu\u4e0a\u5b89\u88c5paddlepaddl":10,"\u5185\u5b58\u4e0d\u591f\u7528\u7684\u60c5\u51b5":23,"\u5377\u79ef\u6a21\u578b":13,"\u53c2\u8003":23,"\u53cc\u5c42rnn\u4ecb\u7ecd":2,"\u53cc\u5c42rnn\u7684\u4f7f\u7528":2,"\u53cc\u5c42rnn\u914d\u7f6e\u4e0e\u793a\u4f8b":1,"\u53cc\u8fdb\u53cc\u51fa":1,"\u53ef\u80fd\u7684\u5185\u5b58\u6cc4\u9732\u95ee\u9898":23,"\u53ef\u80fd\u9047\u5230\u7684\u95ee\u9898":10,"\u547d\u4ee4\u884c\u53c2\u6570":[13,16,25],"\u548c":0,"\u56fe\u50cf":12,"\u57fa\u672c\u539f\u7406":2,"\u5b89\u88c5":[8,13],"\u5b89\u88c5\u7f16\u8bd1paddlepaddle\u9700\u8981\u7684\u4f9d\u8d56":6,"\u5b89\u88c5paddlepaddle\u7684docker\u955c\u50cf":9,"\u5e38\u7528\u6a21\u578b":12,"\u5e8f\u5217\u6a21\u578b\u6570\u636e\u63d0\u4f9b":23,"\u5f00\u53d1\u6307\u5357":14,"\u6027\u80fd\u95ee\u9898":9,"\u603b\u4f53\u6548\u679c\u603b\u7ed3":13,"\u63a8\u8350":12,"\u652f\u6301\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165\u7684layer":0,"\u6570\u636e\u5411\u6a21\u578b\u4f20\u9001":13,"\u6570\u636e\u63d0\u4f9b":25,"\u6570\u636e\u683c\u5f0f\u51c6\u5907":13,"\u65f6\u5e8f\u6a21\u578b":13,"\u6784\u5efapaddlepaddl":3,"\u6982\u8ff0":[0,2],"\u6a21\u578b\u4e2d\u7684\u914d\u7f6e":1,"\u6ce8\u610f\u4e8b\u9879":[9,23],"\u751f\u6210\u6d41\u7a0b\u7684\u4f7f\u7528\u65b9\u6cd5":2,"\u7528\u6237\u63a5\u53e3":25,"\u76f8\u5173\u6982\u5ff5":2,"\u793a\u4f8b1":1,"\u793a\u4f8b2":1,"\u793a\u4f8b3":1,"\u793a\u4f8b4":1,"\u7b80\u5355\u7684\u4f7f\u7528\u573a\u666f":23,"\u7b97\u6cd5\u6559\u7a0b":14,"\u7f16\u8bd1":8,"\u7f16\u8bd1\u4e0e\u5b89\u88c5":8,"\u7f51\u7edc\u7ed3\u6784":13,"\u81ea\u5b9a\u4e49\u4e00\u4e2adataprovid":24,"\u81ea\u7136\u8bed\u8a00\u5904\u7406":12,"\u8bad\u7ec3\u6a21\u578b":13,"\u8bad\u7ec3\u6d41\u7a0b\u7684\u4f7f\u7528\u65b9\u6cd5":2,"\u8bbe\u7f6epaddlepaddle\u7684\u7f16\u8bd1\u9009\u9879":4,"\u8bcd\u5411\u91cf\u6a21\u578b":13,"\u8bfb\u53d6\u53cc\u5c42\u5e8f\u5217\u7684\u65b9\u6cd5":1,"\u8f93\u5165":2,"\u8f93\u5165\u4e0d\u7b49\u957f":1,"\u8f93\u5165\u793a\u4f8b":2,"\u8f93\u51fa":2,"\u8f93\u51fa\u65e5\u5fd7":13,"\u8fdc\u7a0b\u8bbf\u95ee\u95ee\u9898\u548c\u4e8c\u6b21\u5f00\u53d1":9,"\u903b\u8f91\u56de\u5f52\u6a21\u578b":13,"\u914d\u7f6e\u4e2d\u7684\u6570\u636e\u52a0\u8f7d\u5b9a\u4e49":13,"\u9644\u5f55":13,"\u96c6\u7fa4\u8bad\u7ec3":11,"\u9884\u6d4b":[13,25],"beam_search\u7684\u751f\u6210":1,"blas\u76f8\u5173\u7684\u7f16\u8bd1\u9009\u9879":4,"bool\u578b\u7684\u7f16\u8bd1\u9009\u9879":4,"config\u6587\u4ef6\u627e\u4e0d\u5230":10,"cudnn\u76f8\u5173\u7684\u7f16\u8bd1\u9009\u9879":4,"driver\u627e\u4e0d\u5230":10,"group\u6559\u7a0b":2,"make\u548cmak":7,"paddlepaddle\u5feb\u901f\u5165\u95e8\u6559\u7a0b":13,"paddlepaddle\u63d0\u4f9b\u7684docker\u955c\u50cf\u7248\u672c":9,"paddlepaddle\u6587\u6863":14,"paddlepaddle\u7684\u6570\u636e\u63d0\u4f9b":22,"paddlepaddle\u7684\u7f16\u8bd1\u9009\u9879":4,"paddlepaddle\u7684bool\u578b\u7f16\u8bd1\u9009\u9879":4,"paddlepaddle\u7684cblas\u7f16\u8bd1\u9009\u9879":4,"paddlepaddle\u7684python\u9884\u6d4b\u63a5\u53e3":26,"pserver\u7684\u547d\u4ee4\u884c\u53c2\u6570":19,"pydataprovider2\u7684\u4f7f\u7528":23,"python\u6570\u636e\u52a0\u8f7d\u811a\u672c":13,"so\u627e\u4e0d\u5230":10,"subseq\u95f4\u65e0memori":1,"subseq\u95f4\u6709memori":1,"train\u7684\u547d\u4ee4\u884c\u53c2\u6570":20,"version\u7684\u547d\u4ee4\u884c\u53c2\u6570":21,algorithm:13,appendix:13,architectur:13,argument:13,cach:23,command:13,configur:13,convolut:13,cuda:[4,10],data:13,dataprovid:22,docker:3,expand_lay:0,first_seq:0,image:3,init_hook:23,input_typ:23,instal:7,install:13,last_seq:0,libcudart:10,libcudnn:10,line:13,log:13,logist:13,memori:2,model:13,network:13,optimiz:13,overview:13,paddl:[19,20,21],pooling_lay:0,predict:13,prepar:13,provid:[13,23],recurr:2,refer:23,regress:13,script:13,sequenc:13,summari:13,time:13,train:13,transfer:13,vector:13,word:13}})
\ No newline at end of file
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或