Of Stack

PyTorch Checking GPU Version Installed Successfully

  • 2021-10-11 19:11:03
  • OfStack

Check at the anaconda command line: 


(base) PS C:\Users\chenxuqi> conda deactivate
PS C:\Users\chenxuqi> conda activate ssd
(ssd) PS C:\Users\chenxuqi> python
Python 3.6.12 |Anaconda, Inc.| (default, Sep 9 2020, 00:29:25) [MSC v.1916 64 bit (AMD64)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import torch
>>> print(torch.__version__)
0.4.0
>>> print(torch.cuda.is_available())
True
>>>
>>>
>>>

Check the cuda installation from the cmd command line: 


Microsoft Windows [ Version  10.0.18363.1139]
(c) 2019 Microsoft Corporation . All rights reserved. 

C:\Users\chenxuqi>nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2017 NVIDIA Corporation
Built on Fri_Sep__1_21:08:32_Central_Daylight_Time_2017
Cuda compilation tools, release 9.0, V9.0.176

C:\Users\chenxuqi>
C:\Users\chenxuqi>

Execute code using GPU: 


import time
import torch 

##################################################

for i in range(1,10):
  start = time.time()
  a = torch.FloatTensor(i*100,1000,1000)
  a = a.cuda() #a = a
  a = torch.matmul(a,a)
  end = time.time() - start
  print(end)

Implementation results:

Note that the video memory here is too small and overflowed... but the installation was successful... 


Windows PowerShell

 Try a new cross-platform  PowerShell https://aka.ms/pscore6

PS C:\Users\chenxuqi\Desktop\ New Folder > & 'D:\Anaconda3\envs\ssd\python.exe' 'c:\Users\chenxuqi\.vscode\extensions\ms-python.python-2020.10.332292344\pythonFiles\lib\python\debugpy\launcher' '50571' '--' 'c:\Users\chenxuqi\Desktop\ New Folder \testGPU.py'
3.6260359287261963
0.6305170059204102
0.9055967330932617
1.3199987411499023
1.5979139804840088
2.0483360290527344
THCudaCheck FAIL file=c:\programdata\miniconda3\conda-bld\pytorch_1524549877902\work\aten\src\thc\generic/THCStorage.cu line=58 error=2 : out of memory
Traceback (most recent call last):
 File "c:\Users\chenxuqi\Desktop\ New Folder \testGPU.py", line 10, in <module>
  a = torch.matmul(a,a)
RuntimeError: cuda runtime error (2) : out of memory at c:\programdata\miniconda3\conda-bld\pytorch_1524549877902\work\aten\src\thc\generic/THCStorage.cu:58
PS C:\Users\chenxuqi\Desktop\ New Folder > conda activate ssd
PS C:\Users\chenxuqi\Desktop\ New Folder >

Supplement: Offline installation of pytorch to verify successful installation of gpu version

Installing pytorch online with conda command will interrupt the download. Add pip Tsinghua University image source https://pypi. tuna. tsinghua. edu. cn/simple/torch, install the package, and then use pip command 


pip install " The path of the installation package under "

For example: 


pip install "C:\Users\28614\Desktop\pytorch-nightly-cpu-1.0.0.dev20181222-py3.7_cpu_0.tar.bz2"

To install gpu version, you need to install cuda and cudnn.

Verify that gpu version is installed successfully 


import torch
print(torch.cuda.is_available())

If True is returned, the installation is successful

Related articles: