Why do I get a segmentation fault for memory checking?

1

I am using Python 1.7

In [1]: import torch
   ...: from torchvision.models import vgg19
   ...: 
   ...: device = torch.device("cuda:0")

In [2]: 

In [2]: memory = torch.cuda.memory_allocated(device)
Segmentation fault (core dumped)

And my GPU info:

~# nvidia-smi
Fri May 21 13:13:27 2021       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 418.87.01    Driver Version: 418.87.01    CUDA Version: 10.1     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|===============================+======================+======================|
|   0  Tesla V100-SXM2...  On   | 00000000:00:10.0 Off |                    0 |
| N/A   33C    P0    56W / 300W |  15323MiB / 16130MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+
|   1  Tesla V100-SXM2...  On   | 00000000:00:11.0 Off |                    0 |
| N/A   33C    P0    57W / 300W |    462MiB / 16130MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+
|   2  Tesla V100-SXM2...  On   | 00000000:00:12.0 Off |                    0 |
| N/A   39C    P0    56W / 300W |    464MiB / 16130MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+
|   3  Tesla V100-SXM2...  On   | 00000000:00:13.0 Off |                    0 |
| N/A   55C    P0   283W / 300W |  15965MiB / 16130MiB |     99%      Default |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                       GPU Memory |
|  GPU       PID   Type   Process name                             Usage      |
|=============================================================================|
+-----------------------------------------------------------------------------+

However, if I check on terminal, it has no segmentation fault:

# python -c 'import torch; print(torch.cuda.memory_allocated(torch.cuda.device("cuda:0")))'
0

Also, the GPU is used fully, but why this check also returns 0?

# python -c 'import torch; print(torch.cuda.memory_allocated(torch.cuda.device("cuda:3")))'
0
2

The code you provide looks fine; I run it just now, and there is no error occurred. If possible, you can launch a new terminal to try again or provide more detailed information about your exp env, like torch version.