问题现象:
之前安装后不能在添加pytorch和paddlepaddle框架
![](https://file.jishuzhan.net/article/1749316417917816834/27e3d966bdca84f9248804f15c579a2b.webp)
原因(疑似):
在终端中显示pytorch和paddle在C盘但是安装是安装在J盘
解决办法:
卸载、删除文件重新安装后可以看到文件位置在J盘中
![](https://file.jishuzhan.net/article/1749316417917816834/4a5c2dd2b63a882ace2bbea2b8862a97.webp)
但是选择时还是显示Conda executable is not found
![](https://file.jishuzhan.net/article/1749316417917816834/512c8e2f5d65c9ec23f530a86faa57b5.webp)
需要先在你自己下载Anaconda的路径下找到Script目录下的conda.exe,然后双击,就会出来你事先创建的虚拟环境
![](https://file.jishuzhan.net/article/1749316417917816834/c1b2de49acbebaddb7d62afe2f8861ca.webp)
![](https://file.jishuzhan.net/article/1749316417917816834/b9bc7c18469a092962b3818e1074ebf1.webp)
![](https://file.jishuzhan.net/article/1749316417917816834/a0e28c74de52b6821c5a13e1c69e66f8.webp)
![](https://file.jishuzhan.net/article/1749316417917816834/81cf3de727c36d21cbd1ec9a2fec1499.webp)
我们在创建的工程里面创建一个python脚本,在脚本中运行如下代码,查看是否anconda在安装pytorch环境的时候也安装了cuda和cudnn。
python
import torch
print(torch.cuda.is_available())
print(torch.backends.cudnn.is_available())
print(torch.cuda_version)
print(torch.backends.cudnn.version())
![](https://file.jishuzhan.net/article/1749316417917816834/847e5687a34e11dd936d70989b71f7fe.webp)
python
import paddle
print(paddle.utils.run_check())
![](https://file.jishuzhan.net/article/1749316417917816834/7fc04aeb544f394d3177406f2f64ce51.webp)