转于：https://ptorch.com/news/145.html
pytorch v0.3.1发布，添加余弦退火方案来调整学习率以及添加random_split/reduce等函数，其实更新有一段时间了，很多小伙伴安装的时候发现被墙了无法使用，所以我整理了一下方便大家使用。你可以参考自己的服务器和python版本进行选择：
一、使用conda安装pytorch 0.3.1和torchvision
| 包管理器 | 服务器 | CUDA | Python版本 | 操作方法 |
|---|---|---|---|---|
| conda | linux | cuda8 | python2.7 | conda install pytorch torchvision -c pytorch |
| conda | linux | cuda9.0 | python2.7 | conda install pytorch torchvision cuda90 -c pytorch |
| conda | linux | cuda9.1 | python2.7 | conda install pytorch torchvision cuda91 -c pytorch |
| conda | linux | cudanone | python2.7 | conda install pytorch-cpu torchvision -c pytorch |
| conda | linux | cuda8 | python3.5 | conda install pytorch torchvision -c pytorch |
| conda | linux | cuda9.0 | python3.5 | conda install pytorch torchvision cuda90 -c pytorch |
| conda | linux | cuda9.1 | python3.5 | conda install pytorch torchvision cuda91 -c pytorch |
| conda | linux | cudanone | python3.5 | conda install pytorch-cpu torchvision -c pytorch |
| conda | linux | cuda8 | python3.6 | conda install pytorch torchvision -c pytorch |
| conda | linux | cuda9.0 | python3.6 | conda install pytorch torchvision cuda90 -c pytorch |
| conda | linux | cuda9.1 | python3.6 | conda install pytorch torchvision cuda91 -c pytorch |
| conda | linux | cudanone | python3.6 | conda install pytorch-cpu torchvision -c pytorch |
| conda | osx | cuda8 | python2.7 | conda install pytorch torchvision -c pytorch # macOS Binaries dont support CUDA install from source if CUDA is needed |
| conda | osx | cuda9.0 | python2.7 | conda install pytorch torchvision -c pytorch # macOS Binaries dont support CUDA install from source if CUDA is needed |
| conda | osx | cuda9.1 | python2.7 | conda install pytorch torchvision -c pytorch # macOS Binaries dont support CUDA install from source if CUDA is needed |
| conda | osx | cudanone | python2.7 | conda install pytorch torchvision -c pytorch # macOS Binaries dont support CUDA install from source if CUDA is needed |
| conda | osx | cuda8 | python3.5 | conda install pytorch torchvision -c pytorch # macOS Binaries dont support CUDA install from source if CUDA is needed |
| conda | osx | cuda9.0 | python3.5 | conda install pytorch torchvision -c pytorch # macOS Binaries dont support CUDA install from source if CUDA is needed |
| conda | osx | cuda9.1 | python3.5 | conda install pytorch torchvision -c pytorch # macOS Binaries dont support CUDA install from source if CUDA is needed |
| conda | osx | cudanone | python3.5 | conda install pytorch torchvision -c pytorch # macOS Binaries dont support CUDA install from source if CUDA is needed |
| conda | osx | cuda8 | python3.6 | conda install pytorch torchvision -c pytorch # macOS Binaries dont support CUDA install from source if CUDA is needed |
| conda | osx | cuda9.0 | python3.6 | conda install pytorch torchvision -c pytorch # macOS Binaries dont support CUDA install from source if CUDA is needed |
| conda | osx | cuda9.1 | python3.6 | conda install pytorch torchvision -c pytorch # macOS Binaries dont support CUDA install from source if CUDA is needed |
| conda | osx | cudanone | python3.6 | conda install pytorch torchvision -c pytorch # macOS Binaries dont support CUDA install from source if CUDA is needed |
二、使用PIP与pip3安装pytorch 0.3.1和torchvision
除了使用conda以外，我们还可以使用pip管理工具，当我们python未python3.x时候，可以使用pip3安装：
| 包管理器 | 服务器 | CUDA | Python版本 | 操作方法 |
|---|---|---|---|---|
| pip | osx | cuda8 | python2.7 | pip install http://download.pytorch.org/whl/torch-0.3.1-cp27-none-macosx_10_6_x86_64.whl pip install torchvision # macOS Binaries dont support CUDA install from source if CUDA is needed |
| pip | osx | cuda9.0 | python2.7 | pip install http://download.pytorch.org/whl/torch-0.3.1-cp27-none-macosx_10_6_x86_64.whl pip install torchvision # macOS Binaries dont support CUDA install from source if CUDA is needed |
| pip | osx | cuda9.1 | python2.7 | pip install http://download.pytorch.org/whl/torch-0.3.1-cp27-none-macosx_10_6_x86_64.whl pip install torchvision # macOS Binaries dont support CUDA install from source if CUDA is needed |
| pip | osx | cudanone | python2.7 | pip install http://download.pytorch.org/whl/torch-0.3.1-cp27-none-macosx_10_6_x86_64.whl pip install torchvision # macOS Binaries dont support CUDA install from source if CUDA is needed |
| pip | osx | cuda8 | python3.5 | pip3 install http://download.pytorch.org/whl/torch-0.3.1-cp35-cp35m-macosx_10_6_x86_64.whl pip3 install torchvision # macOS Binaries dont support CUDA install from source if CUDA is needed |
| pip | osx | cuda9.0 | python3.5 | pip3 install http://download.pytorch.org/whl/torch-0.3.1-cp35-cp35m-macosx_10_6_x86_64.whl pip3 install torchvision # macOS Binaries dont support CUDA install from source if CUDA is needed |
| pip | osx | cuda9.1 | python3.5 | pip3 install http://download.pytorch.org/whl/torch-0.3.1-cp35-cp35m-macosx_10_6_x86_64.whl pip3 install torchvision # macOS Binaries dont support CUDA install from source if CUDA is needed |
| pip | osx | cudanone | python3.5 | pip3 install http://download.pytorch.org/whl/torch-0.3.1-cp35-cp35m-macosx_10_6_x86_64.whl pip3 install torchvision # macOS Binaries dont support CUDA install from source if CUDA is needed |
| pip | osx | cuda8 | python3.6 | pip3 install http://download.pytorch.org/whl/torch-0.3.1-cp36-cp36m-macosx_10_7_x86_64.whl pip3 install torchvision # macOS Binaries dont support CUDA install from source if CUDA is needed |
| pip | osx | cuda9.0 | python3.6 | pip3 install http://download.pytorch.org/whl/torch-0.3.1-cp36-cp36m-macosx_10_7_x86_64.whl pip3 install torchvision # macOS Binaries dont support CUDA install from source if CUDA is needed |
| pip | osx | cuda9.1 | python3.6 | pip3 install http://download.pytorch.org/whl/torch-0.3.1-cp36-cp36m-macosx_10_7_x86_64.whl pip3 install torchvision # macOS Binaries dont support CUDA install from source if CUDA is needed |
| pip | osx | cudanone | python3.6 | pip3 install http://download.pytorch.org/whl/torch-0.3.1-cp36-cp36m-macosx_10_7_x86_64.whl pip3 install torchvision # macOS Binaries dont support CUDA install from source if CUDA is needed |
| pip | linux | cudanone | python2.7 | pip install http://download.pytorch.org/whl/cpu/torch-0.3.1-cp27-cp27mu-linux_x86_64.whl pip install torchvision # if the above command does not work then you have python 2.7 UCS2 use this command pip install http://download.pytorch.org/whl/cpu/torch-0.3.1-cp27-cp27m-linux_x86_64.whl |
| pip | linux | cuda8 | python2.7 | pip install http://download.pytorch.org/whl/cu80/torch-0.3.1-cp27-cp27mu-linux_x86_64.whl pip install torchvision # if the above command does not\ work then you have python 2.7 UCS2 use this command pip install http://download.pytorch.org/whl/cu80/torch-0.3.1-cp27-cp27m-linux_x86_64.whl |
| pip | linux | cuda9.0 | python2.7 | pip install http://download.pytorch.org/whl/cu90/torch-0.3.1-cp27-cp27mu-linux_x86_64.whl pip install torchvision # if the above command does not\ work then you have python 2.7 UCS2 use this command pip install http://download.pytorch.org/whl/cu90/torch-0.3.1-cp27-cp27m-linux_x86_64.whl |
| pip | linux | cuda9.1 | python2.7 | pip install http://download.pytorch.org/whl/cu91/torch-0.3.1-cp27-cp27mu-linux_x86_64.whl pip install torchvision # if the above command does not\ work then you have python 2.7 UCS2 use this command pip install http://download.pytorch.org/whl/cu91/torch-0.3.1-cp27-cp27m-linux_x86_64.whl |
| pip | linux | cudanone | python3.5 | pip3 install http://download.pytorch.org/whl/cpu/torch-0.3.1-cp35-cp35m-linux_x86_64.whl pip3 install torchvision |
| pip | linux | cuda8 | python3.5 | pip3 install http://download.pytorch.org/whl/cu80/torch-0.3.1-cp35-cp35m-linux_x86_64.whl pip3 install torchvision |
| pip | linux | cuda9.0 | python3.5 | pip3 install http://download.pytorch.org/whl/cu90/torch-0.3.1-cp35-cp35m-linux_x86_64.whl pip3 install torchvision |
| pip | linux | cuda9.1 | python3.5 | pip3 install http://download.pytorch.org/whl/cu91/torch-0.3.1-cp35-cp35m-linux_x86_64.whl pip3 install torchvision |
| pip | linux | cudanone | python3.6 | pip3 install http://download.pytorch.org/whl/cpu/torch-0.3.1-cp36-cp36m-linux_x86_64.whl pip3 install torchvision |
| pip | linux | cuda8 | python3.6 | pip3 install http://download.pytorch.org/whl/cu80/torch-0.3.1-cp36-cp36m-linux_x86_64.whl pip3 install torchvision |
| pip | linux | cuda9.0 | python3.6 | pip3 install http://download.pytorch.org/whl/cu90/torch-0.3.1-cp36-cp36m-linux_x86_64.whl pip3 install torchvision |
| pip | linux | cuda9.1 | python3.6 | pip3 install http://download.pytorch.org/whl/cu91/torch-0.3.1-cp36-cp36m-linux_x86_64.whl pip3 install torchvision |
3. 安装pytorch
用pip install .whl文件即可
pip install torch-0.3.0.post4-cp27-cp27mu-linux_x86_64.whl
pip install torch vision
有了镜像加持和.whl文件的本地化,安装很顺利
torch vision安装非常迅速
测试
在终端中按以下顺序运行命令:
python在打开的python命令行中运行
import torch不报错的话即代表安装成功.
其实安装方法都大同小异，如果你想要查看pytorch 0.3.1的更新内容的话可以查看pytorch v0.3.1发布，添加余弦退火方案来调整学习率以及添加random_split/reduce等函数如果你已经安装了pytorch的话可以测试Pytorch如何更新版本与卸载，使用pip,conda更新卸载Pytorch