I’m having a rather unusual problem importing PyTorch. I use VirtualBox 6.1 and Ubuntu 20.4 to run my ROS application with PyTorch as backend. Since yesterday I have the following error when I want to train my neural network. It worked a couple of weeks ago and I haven’t changed the code sind then. the only thing changed was the PyTorch 1.9.0 release.
Traceback (most recent call last): File "/home/marta/catkin_ws/src/imagineer/src/ai_service/ai_service.py", line 2, in <module> import torch, os, numpy ImportError: No module named torch
PyTorch is definitely installed I tested it with:
Python 3.8.5 (default, May 27 2021, 13:30:53) [GCC 9.3.0] on linux Type "help", "copyright", "credits" or "license" for more information. >>> import torch >>> print(torch.__version__) 1.8.0+cpu >>> torch.cuda.is_available() False >>>
I already deleted my
catkin_ws Workspace, deleted and reinstalled ROS, reinstalled PyTorch 1.9.0 and 1.8.0 (current version now is 1.8.0) but the error message remains. Any idea where the problem could be?