python3.8安装torch1.8.0GPU版
- torch官网查看对应版本安装命令 ,https://pytorch.org/get-started/previous-versions/
选择CUDA11.1,进行安装2.运行代码报错,卸载torch后,重新安装,torch对用版本安装包链接:https://download.pytorch.org/whl/torch_stable.html
pip uninstall torch
# CUDA 11.1
pip install torch==1.8.0+cu111 torchvision==0.9.0+cu111 -f https://download.pytorch.org/whl/torch_stable.html
下载过慢,使用python3.8+torch1.9.0+cuda11.1环境运行代码成功
代码要求环境为python3.8+torch1.8.0,因此可能torch版本1.8和1.9相差不大