总结：pytorch对应版本安装

用户 LifeIsArtOfWar 补充了下载链接: Links for torch ，可以通过文件名查是否存在对应版本。

pytorch 与 torchvision 的对应关系 可以不用考虑，pytorch 官网安装命令的 pytorch 与 torchvision 几乎是一对一关系

pytorch 和 cudatoolkit 对应版本

cuda 和 cudatoolkit 版本版本命名相同 （1表示官网有安装命令，空未知）

cudatoolkit	torch 1.5	torch 1.6	torch 1.7	torch 1.8	torch 1.9	torch 1.10	torch 1.11
9.2	1	1	1
10.1	1	1	1	1
10.2	1	1	1	1	1	1	1
11.0			1
11.1				1	1	1
11.3				1	1	1	1

cudatoolkit	9.2	10.1	10.2	11.0	11.1	11.3
torch 1.5	1	1	1
torch 1.6	1	1	1
torch 1.7	1	1	1	1
torch 1.8		1	1		1	1
torch 1.9			1		1	1
torch 1.10			1		1	1
torch 1.11			1			1

pytorch 和 python 对应版本

发现python与torch随意组合会出问题 （1则可行，0出过问题且未尝试解决，空未知）

python	3.6	3.7	3.8	3.9	3.10
torch 1.5	1
torch 1.6		1	1
torch 1.7		1	1	1
torch 1.8			1	1
torch 1.9			0	1
torch 1.10				1	0
torch 1.11				1	1
torch 1.12				1	1

pytorch 和 python 发布时间

python2 停更了

时间	python3	torch
2018-06-27	3.7
2019-10-14	3.8
2020-10-05	3.9
2020-10-28		1.7
2021-06-16		1.9
2021-10-04	3.10
2021-10-21		1.10
2022-03-11		1.11
2023-06-27	3.7 End
2024-10	3.8 End
2025-10	3.9 End
2026-10	3.10 End

pytorch15

conda create -n py36_torch15_cuda92 python=3.6
conda activate py36_torch15_cuda92
conda install pytorch==1.5.1 torchvision==0.6.1 cudatoolkit=9.2 -c pytorch

conda create -n py36_torch15_cuda101 python=3.6
conda activate py36_torch15_cuda101
conda install pytorch==1.5.1 torchvision==0.6.1 cudatoolkit=10.1 -c pytorch

conda create -n py36_torch15_cuda102 python=3.6
conda activate py36_torch15_cuda102
conda install pytorch==1.5.1 torchvision==0.6.1 cudatoolkit=10.2 -c pytorch

pytorch16

conda create -n py38_torch16_cuda92 python=3.8
conda activate py38_torch16_cuda92
conda install pytorch==1.6.0 torchvision==0.7.0 cudatoolkit=9.2 -c pytorch

conda create -n py38_torch16_cuda101 python=3.8
conda activate py38_torch16_cuda101
conda install pytorch==1.6.0 torchvision==0.7.0 cudatoolkit=10.1 -c pytorch

conda create -n py38_torch16_cuda102 python=3.8
conda activate py38_torch16_cuda102
conda install pytorch==1.6.0 torchvision==0.7.0 cudatoolkit=10.2 -c pytorch

pytorch17

conda create -n py38_torch17_cuda92 python=3.8
conda activate py38_torch17_cuda92
conda install pytorch==1.7.1 torchvision==0.8.2 torchaudio==0.7.2 cudatoolkit=9.2 -c pytorch

conda create -n py38_torch17_cuda101 python=3.8
conda activate py38_torch17_cuda101
conda install pytorch==1.7.1 torchvision==0.8.2 torchaudio==0.7.2 cudatoolkit=10.1 -c pytorch

conda create -n py38_torch17_cuda102 python=3.8
conda activate py38_torch17_cuda102
conda install pytorch==1.7.1 torchvision==0.8.2 torchaudio==0.7.2 cudatoolkit=10.2 -c pytorch

conda create -n py38_torch17_cuda110 python=3.8
conda activate py38_torch17_cuda110
conda install pytorch==1.7.1 torchvision==0.8.2 torchaudio==0.7.2 cudatoolkit=11.0 -c pytorch

pytorch18

conda create -n py39_torch18_cuda102 python=3.9
conda activate py39_torch18_cuda102
conda install pytorch==1.8.1 torchvision==0.9.1 torchaudio==0.8.1 cudatoolkit=10.2 -c pytorch

conda create -n py39_torch18_cuda113 python=3.9
conda activate py39_torch18_cuda113
conda install pytorch==1.8.1 torchvision==0.9.1 torchaudio==0.8.1 cudatoolkit=11.3 -c pytorch -c conda-forge

pytorch19

conda create -n py39_torch19_cuda102 python=3.9
conda activate py39_torch19_cuda102
conda install pytorch==1.9.1 torchvision==0.10.1 torchaudio==0.9.1 cudatoolkit=10.2 -c pytorch

conda create -n py39_torch19_cuda113 python=3.9
conda activate py39_torch19_cuda113
conda install pytorch==1.9.1 torchvision==0.10.1 torchaudio==0.9.1 cudatoolkit=11.3 -c pytorch -c conda-forge

pytorch110

conda create -n py39_torch110_cuda102 python=3.9
conda activate py39_torch110_cuda102
conda install pytorch==1.10.1 torchvision==0.11.2 torchaudio==0.10.1 cudatoolkit=10.2 -c pytorch

conda create -n py39_torch110_cuda113 python=3.9
conda activate py39_torch110_cuda113
conda install pytorch==1.10.1 torchvision==0.11.2 torchaudio==0.10.1 cudatoolkit=11.3 -c pytorch -c conda-forge