How to fix segmentation fault in pytorch without GPU on RHEL6?

Question

linux

Actually the problem is that it is impossible to run torch on a virtual machine without a GPU.
* python 3.6
* Installed glibc-2.14, registered in LD_LIBRARY_PATH
* Installed pytorch-cpu for Linux via pip3 (pip3 install https://download.pytorch.org/whl/cpu/torch-1.0.1-c...
* Installed torchvision via pip3 (pip3 install torchvision)
$ python3 -c "import pytorch"
segmentation fault...
Google doesn't help
RHEL6 problem and need to try on RHEL7?
if the drivers are not installed, maybe it cannot even make a request to the video and therefore it crashes?)
A separate problem is that there is no way to install gdd.
Can someone confirm that on a virtual machine without CUDA and without a GPU on RHEL6 torch works?