为MMclassification配置环境“pytorch-gpu-classification”
下载安装CUDA,cudnn,查看电脑显卡等请参考:
参考:https://blog.csdn.net/qq_43855258/article/details/128690570
参考:(我的)https://blog.csdn.net/tjnkyqcy/article/details/127692763
打开Anaconda的命令行界面控制端Anaconda Promp,输入以下指令,建立新环境
conda create -n 【name_of_new_environment】python=【python_version】
经过反复实验,最终合理的配置如下:
0,python3.8.16
1,CUDA11.3 最新的cudnn11.x
2,torch1.10 torchvision0.11.1 torchaudio0.10.0
torch-1.10.0+cu113-cp38-cp38-win_amd64.whl
torchaudio-0.10.0+cu113-cp38-cp38-win_amd64.whl
torchvision-0.11.1+cu113-cp38-cp38-win_amd64.whl
https://download.pytorch.org/whl/torch_stable.html
3,MMCV1.7.0
mmcv_full-1.7.0-cp38-cp38-win_amd64.whl
链接: https://download.openmmlab.com/mmcv/dist/cu113/torch1.10.0/index.html
4,Xlwings0.30.6
5,matplotlib3.7.1
6,opencv_python-3.4.13.47-cp38-cp38-win_amd64.whl
7,grad_cam1.3.6
8,tensorboard 2.13.0
9,tensorboardX2.6.1
10,setuptools==56.1.0 (该包自动安装会与tensorboard不匹配)