first, from CUDA Toolkit Archive to choose CUDA Toolkit to download.
we choose the version: CUDA Toolkit 12.4.1.
1
2
3
4
5
6
|
❯ nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2024 NVIDIA Corporation
Built on Thu_Mar_28_02:30:10_Pacific_Daylight_Time_2024
Cuda compilation tools, release 12.4, V12.4.131
Build cuda_12.4.r12.4/compiler.34097967_0
|
install cuDNN#
from cuDNN Archive to choose cuDNN to download.
we choose the version: cuDNN 9.1.0.
install pytorch#
Get pip cmd from Installing previous versions of PyTorch
we choose torch==2.4.0:
1
|
pip install torch==2.4.0 torchvision==0.19.0 torchaudio==2.4.0 --index-url https://download.pytorch.org/whl/cu124
|
Check GPU(cuda):
1
2
3
4
|
>>> import torch
>>> torch.cuda.is_available()
True
>>>
|
评论