Other related articles of the author: Pytorch environment detailed installation tutorial [Win10+CUDA upgrade 11.6+cudNN+Anaconda3 virtual environment+pycharm]
Official website: https://pytorch.org/get-started/previous-versions/
CUDA Toolkit version and available PyTorch corresponding relationship summary (refer to the official website)
CUDA Toolkit version | Available PyTorch versions |
---|---|
7.5 | 0.4.1 ,0.3.0, 0.2.0,0.1.12-0.1.6 |
8.0 | 1.1.0,1.0.0 ,0.4.1 |
9.0 | 1.1.0,1.0.1, 1.0.0,0.4.1 |
9.2 | 1.7.1,1.7.0,1.6.0,1.5.1,1.5.0,1.4.0,1.2.0,0.4.1 |
10.0 | 1.2.0,1.1.0,1.0.1 ,1.0.0 |
10.1 | 1.7.1,1.7.0,1.6.0,1.5.1,1.5.0, 1.4.0,1.3.0 |
10.2 | 1.12.1,1.12.0,1.11.0,1.10.1,1.10.0,1.9.1,1.9.0,1.8.1,1.8.0,1.7.1,1.7.0,1.6.0,1.5.1,1.5.0 |
11.0 | 1.7.1,1.7.0 |
11.1 | 1.8.0 |
11.3 | 1.12.1,1.12.0,1.11.0,1.10.1,1.10.0,1.9.1,1.9.0,1.8.1,1.8.0 |
11.6 | 1.13.1,1.13.0,1.12.1,1.12.0 |
11.7 | 1.13.1,1.13.0 |
After the cuda version is determined, pytorch will depend on this cuda version, because the installation of pytroch corresponds to cuda. For example, in the figure below, torch1.11.0 is only suitable for cuda10.2 and 11.3; it is not suitable for cuda10.1 and other versions of cuda.