PyTorch-GPU CUDA cuDnn安装, 入门学习笔记

PyTorch 1.6学习

安装

1.对Anaconda进行一个新的用于pytorch的分区

conda create -n pytorch python=3.6

(base) C:\Users\26330>conda activate pytorch

(pytorch) C:\Users\26330>pip list
Package    Version
---------- -------
pip        21.3.1
setuptools 58.0.4
wheel      0.37.1

2.进行相关环境

C:\Users\26330>conda activate pytorch

3.安装参考链接
链接: link.

C:\Users\26330>nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2016 NVIDIA Corporation
Built on Mon_Jan__9_17:32:33_CST_2017
Cuda compilation tools, release 8.0, V8.0.60

C:\Users\26330>nvidia-smi
Thu Jan 20 14:51:14 2022
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 457.20       Driver Version: 457.20       CUDA Version: 11.1     |
|-------------------------------+----------------------+----------------------+
| GPU  Name            TCC/WDDM | Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  GeForce GTX 1050   WDDM  | 00000000:01:00.0 Off |                  N/A |
| N/A   31C    P8    N/A /  N/A |     75MiB /  4096MiB |      0%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+

+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|  No running processes found                                                 |
+-----------------------------------------------------------------------------+


conda install pytorch==1.1.0 torchvision==0.3.0 cudatoolkit=10.0 -c pytorch

检测是否安装成功

>>> import torch
>>> torch.cuda.is_available()
True
>>>

安装插件

conda install nb_conda
(pytorch) C:\Users\26330>jupyter notebook
[I 16:05:30.079 NotebookApp] [nb_conda_kernels] enabled, 3 kernels found
[I 16:05:31.003 NotebookApp] [nb_conda] enabled
[I 16:05:31.005 NotebookApp] Serving notebooks from local directory: C:\Users\26330

安装成功，很不容易