ubuntu16.04 + anaconda + pytorch（或者caffe2）GPU

1 安装驱动

2 安装CUDA和cuDNN

3 安装anaconda

4 安装pytorch

4.1 安装依赖

export CMAKE_PREFIX_PATH="$(dirname $(which conda))/../" # [anaconda root directory]

# Install basic dependencies
conda install numpy pyyaml mkl mkl-include setuptools cmake cffi typing
conda install -c mingfeima mkldnn

# Add LAPACK support for the GPU
conda install -c pytorch magma-cuda80 # or magma-cuda90 if CUDA 9

注：

export CMAKE_PREFIX_PATH="~/home/lv009/anaconda/bin" # [anaconda root directory] #或者直接添加到环境变量中

conda install -c pytorch magma-cuda80 #若安装不成功，则可以下载安装包，进入到你的环境目录下执行conda install 安装包

4.2 安装

git clone --recursive https://github.com/pytorch/pytorch
cd pytorch
python setup.py install

在编译前conda install lmdb，然后在执行python setup.py install前，执行

export DUSE_LMDB=ON

参考

pytorch官网：https://pytorch.org/get-started/locally/#linux-verification

5 给文件夹权限

sudo chmod 777 文件夹名