anaconda安装mxnet失败解析

背景

MXNet是一款深度学习框架，相比其他框架拥有更快的运算速度以及更小的体积，受到众多机器学习爱好者的青睐。然而，在anaconda安装mxnet时，一些用户会遇到安装失败的情况，本文针对这个问题进行解析。

问题描述

在使用anaconda安装mxnet时，可能会出现以下问题：

1. conda install mxnet时报错：

Preparing transaction: done
Verifying transaction: done
Executing transaction: failed
ERROR conda.core.link:_execute(699): An error occurred while installing package 'mxnet-1.5.0'. LinkError: post-install script failed for package mxnet-1.5.0

2. pip install mxnet时安装失败：

pip install mxnet-cu100 -i https://pypi.tuna.tsinghua.edu.cn/simple

Collecting mxnet-cu100
  Downloading https://pypi.tuna.tsinghua.edu.cn/packages/3e/0c/7aae695fc0b4ae2343caf9e50da19fd3c11c29cac814f321aefee68bef8f/mxnet_cu100-1.6.0-py2.py3-none-manylinux1_x86_64.whl (748.3 MB)
...省略部分...
  Running setup.py install for gluoncv
    ERROR: Command errored out with exit status 1: /home/liuyuchuan/anaconda3/envs/py38/bin/python -u -c 'import sys, setuptools, tokenize; sys.argv[0] = '"'"'/tmp/pip-install-tv8qvopn/gluoncv/setup.py'"'"'; __file__='"'"'/tmp/pip-install-tv8qvopn/gluoncv/setup.py'"'"'; f
ailed = subprocess.check_call(cmd[:-1], cwd=wd, **kwargs)
        subprocess.CalledProcessError: Command '('yum', '--disableplugin=fastestmirror', 'install', '-y', 'pkg-config', 'libopencv-dev', 'libblas-dev', 'liblapack-dev', 'python3-dev', 'libx11-dev', 'libatlas-base-dev', 'libgtk-3-dev', 'libboost-all-dev', 'libjpe
g-dev', 'libpng-dev', 'libtiff-dev')' returned non-zero exit status 1.
        ERROR: Failed building wheel for gluoncv
        ERROR: Could not build wheels for gluon