Module not found 'torch'

1

Hi,

I’ve installed the latest version of pytorch (1.12.1) with cuda 11.3 using the following command in a new conda environment:
conda install pytorch torchvision torchaudio cudatoolkit=11.3 -c pytorch

but when i try to import torch: “import torch” and running it i’m getting Module not found error: “torch”

I’m running on Windows 10, and i’ve tried using python=3.8 and python=3.9 and python=3.7 and still getting this error.

Also i made sure that i have cuda 11.3 installed using nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2021 NVIDIA Corporation
Built on Sun_Mar_21_19:24:09_Pacific_Daylight_Time_2021
Cuda compilation tools, release 11.3, V11.3.58
Build cuda_11.3.r11.3/compiler.29745058_0

Any idea what could be the problem and how to resolve it?

2

Are you sure that you are activating the correct environment when importing torch? Just check if that is the issue.

3

yes, unfortunately i activated the correct environment. i try it from ipython on the console and it doesn’t work :frowning:

4

Capture
Capture1253×771 40.3 KB

I even tried to install it using the pip installation from the official site and still the same problem

maybe it happens because i still have cuda 10.2 installed but in the environment variables in CUDA_PATH i have cuda 11.3

Capture1.PNG
Capture1.PNG1198×929 112 KB

Capture2

5

The issue happens because your environment is not activated on the IPython kernel. You can add the environment to IPython terminal using this. Should do the work.

1 Like