Pytorch: cuDNN error: CUDNN_STATUS_EXECUTION_FAILED

python2与python3发生冲突

1.卸载python2.7

sudo apt-get remove python2.7

127.0.0.1
2.卸载python2.7及其依赖

sudo apt-get remove --auto-remove python2.7

3.消除python2.7

sudo apt-get purge python2.7 or sudo apt-get purge --auto-remove python2.7

python3 和 安装虚拟环境

sudo apt update
sudo apt install python3-dev python3-pip
sudo pip3 install -U virtualenv  # system-wide install

mkdir environment + cd environment

创建虚拟环境

# 创建环境
virtualenv --system-site-packages -p python3 ./tf_py3
# 激活环境
source ./tf_py3/bin/activate  # sh, bash, ksh, or zsh
# 升级pip
pip install --upgrade pip

pip list  # show packages installed within the virtual environment

# 退出环境用 （不退出）
deactivate  # don't exit until you're done using TensorFlow

安装tf

pip install --upgrade tensorflow
pip install numpy matplotlib sklearn pandas jupyter

配置jupyter远程

1.查看主目录下是否有.jupyter文件夹

没有则创建一个

jupyter notebook --generate-config

2. 进入.jupyter

vim jupyter_notebook_config.py
修改如下内容：



c= get_config()
c.NotebookApp.ip = '*'
c.NotebookApp.open_browser = False
c.NotebookApp.port = 6006
c.NotebookApp.allow_remote_access = True
c.NotebookApp.allow_root = True

3. 创建workspace文件夹

mkdir workspace
cd workspace
jupyter notebook

4. token输入上创建密码即可

---

转载请注明来源，欢迎对文章中的引用来源进行考证，欢迎指出任何有错误或不够清晰的表达。可以在下面评论区评论，也可以邮件至 luckiestman0@qq.com