3、tensorflow配置gpu加速
<p><a href="https://blog.csdn.net/louishao/article/details/78827691">https://blog.csdn.net/louishao/article/details/78827691</a>、
CuDA8.0.44，CuDNN至少6.0---win10--python3.5--tensorflow==1.3.0</p>
<pre><code class="language-bash">C:\Users\zcr>conda create -n p36-gpu python=3.6
C:\Users\zcr>activate p36-gpu
(p36-gpu) C:\Users\zcr>pip install tensorflow-gpu==1.3.0 -i https://pypi.tuna.tsinghua.edu.cn/simple
(p36-gpu) C:\Users\zcr>python
>>> import tensorflow as tf
>>> sess = tf.Session()
2018-09-06 16:48:23.976673: W C:\tf_jenkins\home\workspace\rel-win\M\windows-gpu\PY\36\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations.
2018-09-06 16:48:23.989645: W C:\tf_jenkins\home\workspace\rel-win\M\windows-gpu\PY\36\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX2 instructions, but these are available on your machine and could speed up CPU computations.
2018-09-06 16:48:24.709735: I C:\tf_jenkins\home\workspace\rel-win\M\windows-gpu\PY\36\tensorflow\core\common_runtime\gpu\gpu_device.cc:955] Found device 0 with properties:
name: GeForce 840M
major: 5 minor: 0 memoryClockRate (GHz) 1.124
pciBusID 0000:03:00.0
Total memory: 2.00GiB
Free memory: 1.66GiB
2018-09-06 16:48:24.721736: I C:\tf_jenkins\home\workspace\rel-win\M\windows-gpu\PY\36\tensorflow\core\common_runtime\gpu\gpu_device.cc:976] DMA: 0
2018-09-06 16:48:24.725216: I C:\tf_jenkins\home\workspace\rel-win\M\windows-gpu\PY\36\tensorflow\core\common_runtime\gpu\gpu_device.cc:986] 0: Y
2018-09-06 16:48:24.732317: I C:\tf_jenkins\home\workspace\rel-win\M\windows-gpu\PY\36\tensorflow\core\common_runtime\gpu\gpu_device.cc:1045] Creating TensorFlow device (/gpu:0) -> (device: 0, name: GeForce 840M, pci bus id: 0000:03:00.0)
>>> a = tf.constant(2)
>>> b = tf.constant(3)
>>> print(sess.run(a+b))
5</code></pre>
<p>安装tensorflow1.8.0-gpu</p>
<pre><code>C:\Users\zcr>nvcc -V
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2017 NVIDIA Corporation
Built on Fri_Sep__1_21:08:32_Central_Daylight_Time_2017
Cuda compilation tools, release 9.0, V9.0.176
C:\Users\zcr>activate keras
(keras) C:\Users\zcr>python
Python 3.6.2 |Continuum Analytics, Inc.| (default, Jul 20 2017, 12:30:02) [MSC v.1900 64 bit (AMD64)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import tensorflow
>>> import tensorflow as tf
>>> print(tf.__version__)
1.8.0
</code></pre>
<h3>新增说明</h3>
<p>1、安装anaconda、配置anaconda加速、配置pip加速
2、下载cuda_9.0.176_win10.exe
3、下载cudnn-9.0-windows10-x64-v7.4.1.5.zip
4、安装cuda_9.0.176_win10.exe</p>
<pre><code>CUDA Toolkit 9.0 Downloads--安装
Download cuDNN v7.1.2 (Mar 21, 2018), for CUDA 9.0--下载完后解压复制到
C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.0 目录下</code></pre>
<p>5、增加环境变量到path</p>
<pre><code>C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.0\lib\x64</code></pre>
<p>6、</p>
<pre><code>C:\Users\zcr>activate p36
(p36) C:\Users\zcr>pip install tensorflow-gpu==1.11
....
(p36) C:\Users\zcr>python
Python 3.6.2 |Continuum Analytics, Inc.| (default, Jul 20 2017, 12:30:02) [MSC v.1900 64 bit (AMD64)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import tensorflow
>>></code></pre>
<p>7、gpu版成功</p>
<h3>2019 04.01</h3>
<p>下载链接
<a href="https://developer.nvidia.com/cuda-10.0-download-archive">https://developer.nvidia.com/cuda-10.0-download-archive</a></p>
<p><a href="https://developer.nvidia.com/rdp/cudnn-download">https://developer.nvidia.com/rdp/cudnn-download</a></p>