Anaconda安装pytorch

1. 修改源，添加 Anaconda Python 免费仓库

sudo gedit ~/.condarc

修改内容如下

channels:
 1. defaults
show_channel_urls: true
default_channels:
 2. https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main
 3. https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/r
 4. https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/msys2
custom_channels:
  conda-forge: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud
  msys2: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud
  bioconda: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud
  menpo: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud
  pytorch: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud
  simpleitk: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud

清除索引缓存，保证用的是镜像站提供的索引

conda clean -i

2. 打开pytorch网站https://pytorch.org/get-started/locally/，选择自己需要的版本
![在这里插入图片描述](https://img-blog.csdnimg.cn/20210217105504207.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L0dyYWNleWluZw==,size_16,color_FFFFFF,t_70#pic_center)

执行下载命令

```bash
conda install pytorch torchvision torchaudio cudatoolkit=10.1 -c pytorch

等待出现以下内容完成安装

udatoolkit=10.1 -c pytorch
Collecting package metadata (current_repodata.json): done
Solving environment: / 
Warning: 2 possible package resolutions (only showing differing packages):
  - https://mirrors.sjtug.sjtu.edu.cn/anaconda/pkgs/main/linux-64/linux-64::conda-4.9.2-py38h06a4308_0
  - defaults/linux-64::conda-4.9.2-py38h06a4308done

## Package Plan ##

  environment location: /home/xxy/anaconda3

  added / updated specs:
    - cudatoolkit=10.1
    - pytorch
    - torchaudio
    - torchvision


The following packages will be downloaded:

    package                    |            build
    ---------------------------|-----------------
    numpy-1.19.2               |   py38h54aff64_0          22 KB  defaults
    pytorch-1.7.1              |py3.8_cuda10.1.243_cudnn7.6.3_0       553.1 MB  pytorch
    torchvision-0.8.2          |       py38_cu101        17.8 MB  pytorch
    typing_extensions-3.7.4.3  |     pyha847dfd_0          25 KB  defaults
    ------------------------------------------------------------
                                           Total:       570.9 MB

The following NEW packages will be INSTALLED:

  blas               anaconda/pkgs/main/linux-64::blas-1.0-mkl
  cudatoolkit        anaconda/pkgs/main/linux-64::cudatoolkit-10.1.243-h6bb024c_0
  intel-openmp       anaconda/pkgs/main/linux-64::intel-openmp-2020.2-254
  libuv              anaconda/pkgs/main/linux-64::libuv-1.40.0-h7b6447c_0
  mkl                anaconda/pkgs/main/linux-64::mkl-2020.2-256
  mkl-service        anaconda/pkgs/main/linux-64::mkl-service-2.3.0-py38he904b0f_0
  mkl_fft            anaconda/pkgs/main/linux-64::mkl_fft-1.2.0-py38h23d657b_0
  mkl_random         anaconda/pkgs/main/linux-64::mkl_random-1.1.1-py38h0573a6f_0
  ninja              anaconda/pkgs/main/linux-64::ninja-1.10.2-py38hff7bd54_0
  numpy              anaconda/pkgs/main/linux-64::numpy-1.19.2-py38h54aff64_0
  numpy-base         anaconda/pkgs/main/linux-64::numpy-base-1.19.2-py38hfa32c7d_0
  pytorch            pytorch/linux-64::pytorch-1.7.1-py3.8_cuda10.1.243_cudnn7.6.3_0
  torchaudio         pytorch/linux-64::torchaudio-0.7.2-py38
  torchvision        pytorch/linux-64::torchvision-0.8.2-py38_cu101
  typing_extensions  anaconda/pkgs/main/noarch::typing_extensions-3.7.4.3-pyha847dfd_0


Proceed ([y]/n)? y


Downloading and Extracting Packages
typing_extensions-3. | 25 KB     | ######################################################################################################################################################################## | 100% 
pytorch-1.7.1        | 553.1 MB  | ######################################################################################################################################################################## | 100% 
torchvision-0.8.2    | 17.8 MB   | ######################################################################################################################################################################## | 100% 
numpy-1.19.2         | 22 KB     | ######################################################################################################################################################################## | 100% 
Preparing transaction: done
Verifying transaction: done
Executing transaction: done

测试是否安装成功

import torch
import torchvision
x = torch.rand(5, 3)
print(x)
torch.cuda.is_available()

显示如下

Python 3.8.5 (default, Sep  4 2020, 07:30:14) 
[GCC 7.3.0] :: Anaconda, Inc. on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import torch
>>> import torchvision
>>> x = torch.rand(5, 3)
>>> print(x)
tensor([[0.6433, 0.1002, 0.7137],
        [0.5428, 0.3415, 0.5244],
        [0.7444, 0.3307, 0.8274],
        [0.5290, 0.0789, 0.3824],
        [0.1561, 0.3142, 0.7884]])
>>> torch.cuda.is_available()
True
>>>