I try to check if torch have been properly installed by script:
import torch
print(torch.__version__)
and have the result: AttributeError: module 'torch' has no attribute 'version'
I've checked if torch is installed in my venv by:
pip freeze
and it is installed(I tried to install as via conda as via pip as suggested at https://pytorch.org/):
- torch==1.7.1+cu110
- torchaudio==0.7.2
- torchvision==0.8.2+cu110
Environment:
- OS: Ubuntu 18.04
- GPU Geforce RTX 3070
- Driver version 460.39
- Cuda Toolkit V11.0.194
- Cudnn V8.0.5.39
- Python V3.6.9
Do not understand what's the problem
torch.py
this will cause your local script to shadow the module namedtorch
. Instead of importing the pytorch module you are importing your localtorch.py
(/home/evgeniy/cnn/torch.py). – Aurthur