学习记录1：pytorch（GPU）安装

我C盘空间不够，重装Anaconda到D盘了： 

Anaconda 安装及修改环境默认位置-物联沃-IOTWORD物联网https://www.iotword.com/10057.html#:~:text=Anaconda%20%E5%AE%89%E8%A3%85%E5%8F%8A%E4%BF%AE%E6%94%B9%E7%8E%AF%E5%A2%83%E9%BB%98%E8%AE%A4%E4%BD%8D%E7%BD%AE%201%201.%E5%AE%98%E7%BD%91%E4%B8%8B%E8%BD%BD%E5%AE%89%E8%A3%85%E5%8C%85%202%202.%E5%AE%89%E8%A3%85%20%E4%B8%80%E8%B7%AFnext%E5%8D%B3%E5%8F%AF%EF%BC%8C%E5%AE%89%E8%A3%85%E4%BD%8D%E7%BD%AE%E5%8F%AF%E6%94%B9%E5%88%B0D%E7%9B%98%203,%E5%A6%82%E6%9E%9C%E5%AE%89%E8%A3%85%E5%9C%A8D%E7%9B%98%EF%BC%8C%E4%BD%86%E6%98%AF%E9%BB%98%E8%AE%A4%E8%99%9A%E6%8B%9F%E7%8E%AF%E5%A2%83%E4%BC%9A%E5%9C%A8C%E7%9B%98%20%E5%8F%AF%E7%94%A8%E4%BB%A5%E4%B8%8B%E5%91%BD%E4%BB%A4%E8%BF%9B%E8%A1%8C%E6%9F%A5%E7%9C%8B%E8%99%9A%E6%8B%9F%E7%8E%AF%E5%A2%83%E4%BD%8D%E7%BD%AE%20conda%20info%20onda%E7%8E%AF%E5%A2%83%E9%BB%98%E8%AE%A4%E5%AE%89%E8%A3%85%E5%9C%A8%E7%94%A8%E6%88%B7%E7%9B%AE%E5%BD%95C%3A%5CUsers%5Cusername.conda%5Cenvs%E4%B8%8B%EF%BC%8C%E5%A6%82%E6%9E%9C%E9%80%89%E6%8B%A9%E9%BB%98%E8%AE%A4%E8%B7%AF%E5%BE%84%EF%BC%8C%E9%82%A3%E4%B9%88%E4%B9%8B%E5%90%8E%E5%88%9B%E5%BB%BA%E8%99%9A%E6%8B%9F%E7%8E%AF%E5%A2%83%EF%BC%8C%E4%B9%9F%E6%98%AF%E5%AE%89%E8%A3%85%E5%9C%A8%E7%94%A8%E6%88%B7%E7%9B%AE%E5%BD%95%E4%B8%8B%E3%80%82%20%E4%B8%8D%E6%83%B3%E5%8D%A0%E7%94%A8C%E7%9B%98%E7%A9%BA%E9%97%B4%EF%BC%8C%E5%8F%AF%E4%BB%A5%E4%BF%AE%E6%94%B9conda%E8%99%9A%E6%8B%9F%E7%8E%AF%E5%A2%83%E8%B7%AF%E5%BE%84%E3%80%82%20%281%29%E9%A6%96%E5%85%88%EF%BC%8C%E6%89%BE%E5%88%B0%E7%94%A8%E6%88%B7%E7%9B%AE%E5%BD%95%E4%B8%8B%E7%9A%84.condarc%E6%96%87%E4%BB%B6%EF%BC%88C%3A%5CUsers%5Cusername%EF%BC%89%E3%80%82%20

 

 

创建环境，激活环境：

conda create -n env_name python=x.x

activate env_name

Anaconda改变虚拟环境安装路径（必成功）_更改anaconda安装路径_哈仔康康的博客-CSDN博客https://blog.csdn.net/weixin_74034112/article/details/129001618?spm=1001.2101.3001.6661.1&utm_medium=distribute.pc_relevant_t0.none-task-blog-2~default~CTRLIST~Rate-1-129001618-blog-126246310.235%5Ev38%5Epc_relevant_anti_vip&depth_1-utm_source=distribute.pc_relevant_t0.none-task-blog-2~default~CTRLIST~Rate-1-129001618-blog-126246310.235%5Ev38%5Epc_relevant_anti_vip&utm_relevant_index=1

 

cmd命令输入：nvidia-smi

 

 如果等级不够可以进行，nvidia驱动升级：NVIDIA GeForce 驱动程序 - N 卡驱动 | NVIDIA

cuda安装：CUDA Toolkit 11.7 Update 1 Downloads | NVIDIA Developer

Pytorchgua官网：PyTorch

 

pytorch，torchvision与python版本对应关系及安装命令_pytorch python版本_R_X_的博客-CSDN博客https://blog.csdn.net/WOSHIRENXIN/article/details/127415609?utm_medium=distribute.pc_relevant.none-task-blog-2~default~baidujs_baidulandingword~default-0-127415609-blog-127674463.235%5Ev38%5Epc_relevant_anti_vip&spm=1001.2101.3001.4242.1&utm_relevant_index=3我的python是3.11，安装的时候报错不兼容，于是降级到3.10 再安装：

conda install python=3.10

conda install pytorch==2.0.0 torchvision==0.15.0 torchaudio==2.0.0 pytorch-cuda=11.7 -c pytorch -c nvidia

官网下的会比较慢，也可以换源试试（我之前试过清华源但是失败了）

True: