安装备忘录(三)
作者:互联网
****CUDA,cuDNN在 cd /usr/local 下
** cuda version:cuda-8.0
path: /usr/local/cuda-8.0
cuda 版本
cat /usr/local/cuda/version.txt
** cudnn version libcudnn.so.5.1.5
path: /usr/local/cudnn-5.1/lib64/libcudnn.so.5.1.5
tensorflow已安装在用户hss@bj.e.cn的anaconda3下,并未与其他用户共享****** (这应该是之前.78服务器的)
Ubuntu下查看tensorflow版本 1.2.0 (tensorflow1.2.0配cuda8.0、cudnn5.1)
h**@B***-P*C-6S-1**0:~$ python
Python 3.6.4 |Anaconda, Inc.| (default, Jan 16 2018, 18:10:19)
[GCC 7.2.0] on linux
Type “help”, “copyright”, “credits” or “license” for more information.
import tensorflow as tf
tf.version
‘1.2.0’
Ubuntu下查看Keras版本 2.2.4
import keras
Using TensorFlow backend.keras.version
‘2.2.4’
pytorch,Keras已安装********************
****pip install torch --user #安装torch,但是很慢很慢
****pip install --pre keras
****git clone +github上的代码地址 即可将当前项目克隆到服务器上(首先在服务器上新建一个文件夹,用于存放该项目)
linux环境下更新pip版本****
pip install -U pip
windows环境下更新pip版本****
python -m pip install -U pip
python -m pip install --upgrade pip
"libcudnn.so.5 cannot open shared object file: No such file or directory"**
解决办法是将一些文件复制到/usr/local/lib文件夹下:
#注意自己CUDA的版本号!
sudo cp /usr/local/cudnn-5.1/lib64/libcudnn.so /usr/local/lib/libcudnn.so && sudo ldconfig
sudo cp /usr/local/cudnn-5.1/lib64/libcudnn.so.5 /usr/local/lib/libcudnn.so.5 && sudo ldconfig
sudo cp /usr/local/cudnn-5.1/lib64/libcudnn.so.5.1.5 /usr/local/lib/libcudnn.so.5.1.5 && sudo ldconfig
此为找到libcudnn的路径,并复制到 /usr/local/lib/libcudnn*
可参考https://blog.csdn.net/u014696921/article/details/60140264
FutureWarning: Conversion of the second argument of issubdtype from float
to np.floating
is deprecated. In future, it will be treated as np.float64 == np.dtype(float).type
.**********
h5py包内出错 解决办法:对h5py进行更新升级
pip install h5py==2.8.0rc1 亲测没毛病
标签:libcudnn,local,备忘录,usr,install,pip,安装,so.5 来源: https://blog.csdn.net/weixin_38612234/article/details/92202496