使用pip3安装PyTorch与PyG，实现NVIDIA CUDA GPU加速

使用python3的pip3命令安装python依赖库。

# python3 -V
Python 3.12.3
# 
# pip3 -V 
pip 25.0.1 from /root/.pyenv/versions/3.12.3/lib/python3.12/site-packages/pip (python 3.12)

Usage:   
  pip3 install [options] <package> ...
  pip3 install [options] -r <requirements file> ... 
  
  pip3 uninstall [options] <package> ...
  pip3 uninstall [options] -r <requirements file> ... 

GPU驱动和Cuda工具

GPU规格：NVIDIA TU104GL [Tesla T4]
驱动和Cuda版本：NVIDIA-SMI 550.127.08 Driver Version: 550.127.08 CUDA Version: 12.4

# lspci | grep -i nvidia
00:07.0 3D controller: NVIDIA Corporation TU104GL [Tesla T4] (rev a1)
# 
# nvidia-smi
Thu Apr 10 12:13:49 2025       
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 550.127.08             Driver Version: 550.127.08     CUDA Version: 12.4     |
|-----------------------------------------+------------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |
|                                         |                        |               MIG M. |
|=========================================+========================+======================|
|   0  Tesla T4                       On  |   00000000:00:07.0 Off |                    0 |
| N/A   29C    P8              9W /   70W |       1MiB /  15360MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+
                                                                                         
+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU   GI   CI        PID   Type   Process name                              GPU Memory |
|        ID   ID                                                               Usage      |
|=========================================================================================|
|  No running processes found                                                             |
+-----------------------------------------------------------------------------------------+
# 
# nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2024 NVIDIA Corporation
Built on Thu_Mar_28_02:18:24_PDT_2024
Cuda compilation tools, release 12.4, V12.4.131
Build cuda_12.4.r12.4/compiler.34097967_0
# 
# nvidia-smi --query-gpu=name,memory.total,utilization.gpu,temperature.gpu --format=csv
name, memory.total [MiB], utilization.gpu [%], temperature.gpu
Tesla T4, 15360 MiB, 0 %, 29

PyTorch安装

PyTorch, Tensors and Dynamic neural networks in Python with strong GPU acceleration

• https://pytorch.org/
• https://github.com/pytorch/pytorch

pip3 install torch torchvision 
torch-2.6.0

pip3 install -r pytorch/requirements.txt 
pip3 install scikit-learn tqdm gpytorch skorch 

PyG安装

PyG is the ultimate library for Graph Neural Networks
Build graph learning pipelines with ease

• https://pyg.org/
• https://pytorch-geometric.readthedocs.io/en/latest/install/installation.html
• https://github.com/pyg-team/pytorch_geometric

pip3 install torch_geometric 
torch_geometric-2.6.1

# python3 -c "import torch; print(torch.__version__)" 
2.6.0+cu124
# python3 -c "import torch; print(torch.version.cuda)" 
12.4
# python3 -c "import torch; print(torch.cuda.device_count())" 
1

pip3 install pyg_lib torch_scatter torch_sparse torch_cluster torch_spline_conv -f https://data.pyg.org/whl/torch-2.6.0+cu124.html 

Python示例代码依赖

pip3 install -r pyg-examples/requirements.txt 

如何导出Python项目的库依赖文件requirements.txt，自行搜索DeepSeek、ChatGPT或Google/谷歌。