win10+vs2015+cuda10.0+cudnn7.6+anaconda3(python3.7)+tensorflow_gpu1.13.1+pycharm(我的电脑配置:i79700k+1070Ti)
提醒:安装之前先查看自己显卡对应的版本,一般安装过程都大同小异!!!!
(1)下载cuda10.0 cuda10.0下载地址
(2)下载cudnn7.6 cudnn下载地址
(3)下载Git,后面能用到 Git下载地址
(4)下载anaconda3 地址
2.安装
(1)vs2015
(2)cuda+cudnn安装 这儿给出别人弄的,这两个安装比较简单,但是记得把cudnn路径添加到系统变量里面
(3)Git安装按照默认步骤来就行,但是要在系统变量下添加路径,我的是:C:\Program Files\Git\bin
(4)anaconda3安装按照默认路径来,以免出现错误,网上有安装教程,自己找
(5)tensorflow_gpu1.13.1安装:
打开开始菜单中 adaconda3文件夹下Anaconda Prompt 程序,输入:
conda create -n tensorflow-gpu python=3.7
输入y回车,接着输入:
conda activate tensorflow-gpu
pip install tensorflow-gpu==1.13.1
如果出现pip版本老旧,更新pip然后再重新安装tensorflow-gpu,同上
python -m pip install –upgrade pip
按照以上步骤安装应该是可以成功的,本人中间没有出现错误,测试是否安装成功,代码如下
要在该状态下输入python (如果没有先win键+R,输入cmd,再输入activate tensorflow_gpu)
import tensorflow as tf
tf.enable_eager_execution()
print(tf.__version__)
hello = tf.constant("hello tensorflow")
print(hello.numpy())
至此,tensorflow-gpu安装成功
(6)安装imgaug
conda install -c conda-forge shapely
pip install imgaug
(7)安装pycocotools
方法一:
pip install cython
pip install git+https://github.com/philferriere/cocoapi.git#subdirectory=PythonAPI
方法二
git clone https://github.com/pdollar/coco.git
cd coco/PythonAPI
# install pycocotools locally
python setup.py build_ext --inplace
# install pycocotools to the Python site-packages
python setup.py build_ext install
(第二种方法我没有成功)
方法三
pip install pycocotools -i https://pypi.tuna.tsinghua.edu.cn/simple
(8)pycharm安装,环境要添加tensorflow-gpu下的python.exe
参考博客:
https://www.gongyesheji.org/?p=951
https://blog.csdn.net/qq_31119155/article/details/83992500
https://blog.csdn.net/doudou_here/article/details/87855273
https://blog.csdn.net/li528405176/article/details/83154062
https://blog.csdn.net/Xiongchao99/article/details/79778046
https://blog.csdn.net/xiaomifanhxx/article/details/84325174
https://blog.csdn.net/qq_36810544/article/details/83582397