在服务器上装mxnet
先把环境调到合适的python版本: export
export PATH=~/miniconda3/bin:$PATH 然后 python下，显示确认下python版本
hjj@user711:~$ python
Python 3.6.5 |Anaconda, Inc.| (default, Apr 29 2018, 16:14:56)
[GCC 7.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> exit()
hjj@user711:~$ 安装使用pip:
pip install mxnet-cu80太慢了，就直接把连接拷贝到自己电脑下载后再上传，放在自己用户的目录下~\
连接： mxnet-cu80
安装：
pip install ~/mxnet_cu80-1.3.0.post0-py2.py3-none-manylinux1_x86_64.whl
hjj@user711:~$ pip install ~/mxnet_cu80-1.3.0.post0-py2.py3-none-manylinux1_x86_64.whl
Processing ./mxnet_cu80-1.3.0.post0-py2.py3-none-manylinux1_x86_64.whl
Requirement already satisfied: requests<2.19.0,>=2.18.4 in ./miniconda3/lib/python3.6/site-packages (from mxnet-cu80==1.3.0.post0) (2.18.4)
Collecting graphviz<0.9.0,>=0.8.1 (from mxnet-cu80==1.3.0.post0)
Downloading https://files.pythonhosted.org/packages/53/39/4ab213673844e0c004bed8a0781a0721a3f6bb23eb8854ee75c236428892/graphviz-0.8.4-py2.py3-none-any.whl
Requirement already satisfied: numpy<1.15.0,>=1.8.2 in ./miniconda3/lib/python3.6/site-packages (from mxnet-cu80==1.3.0.post0) (1.14.3)
Requirement already satisfied: chardet<3.1.0,>=3.0.2 in ./miniconda3/lib/python3.6/site-packages (from requests<2.19.0,>=2.18.4->mxnet-cu80==1.3.0.post0) (3.0.4)
Requirement already satisfied: idna<2.7,>=2.5 in ./miniconda3/lib/python3.6/site-packages (from requests<2.19.0,>=2.18.4->mxnet-cu80==1.3.0.post0) (2.6)
Requirement already satisfied: urllib3<1.23,>=1.21.1 in ./miniconda3/lib/python3.6/site-packages (from requests<2.19.0,>=2.18.4->mxnet-cu80==1.3.0.post0) (1.22)
Requirement already satisfied: certifi>=2017.4.17 in ./miniconda3/lib/python3.6/site-packages (from requests<2.19.0,>=2.18.4->mxnet-cu80==1.3.0.post0) (2018.4.16)
Installing collected packages: graphviz, mxnet-cu80
Successfully installed graphviz-0.8.4 mxnet-cu80-1.3.0.post0
pycharm 下使用
#text.py
import mxnet as mx
from mxnet import nd
from mxnet import autograd
from mxnet import gluon
x = nd.arange(4).reshape((4,1))
print('x:',x)
x.attach_grad()
print('x:',x)
with autograd.record():
y=2*nd.dot(x.T,x)
print('y:',y)
y.backward()
print('x.grad:',x.grad)结果：
x:
[[0.]
[1.]
[2.]
[3.]]
<NDArray 4x1 @cpu(0)>
x:
[[0.]
[1.]
[2.]
[3.]]
<NDArray 4x1 @cpu(0)>
y:
[[28.]]
<NDArray 1x1 @cpu(0)>
x.grad:
[[ 0.]
[ 4.]
[ 8.]
[12.]]
<NDArray 4x1 @cpu(0)>