-
As title. Here's the message.
```
Building wheels for collected packages: torch-scatter
Building wheel for torch-scatter (setup.py) ... error
error: subprocess-exited-with-error
× pytho…
-
thank you for your amazing work. But I met a little problem. when I run python3 gendata.py ./images, I got this
Traceback (most recent call last):
File "gendata.py", line 100, in
device, G,…
-
### 🐛 Describe the bug
When installing via `pip3 install torchvision` I get `torchvision` with CUDA 10.2, yet my systems CUDA is 11.6 while my torch is installed with CUDA 11.3. I managed to fix this…
-
Matplotlib is additionally required to run prepare_cifar10.sh
It would be nice if 'pip install matplotlib' is added to the list.
- conda install pytorch torchvision torchaudio cudatoolkit=10.2 -…
-
### 🐛 Describe the bug
My bug is summarised in the image below. In a nutshell, if you sum along a dimension on a tensor and then convert to cuda, it's different from what is observed by converting to…
-
**Version of Singularity**
What version of Singularity are you using? Run:
```
$ singularity --version
```
singularity-3.7.3
**Describe the bug**
A clear and concise description of what the…
-
### 🐛 Describe the bug
I have an onnx model running a normal endpoint, no issues.
I have a second endpoint that is running a workflow that I think is leaking memory. I see the torch container co…
-
##### System information (version)
- OpenVINO=> 2021.4.689
- Operating System / Platform => Windows 64 Bit
- Compiler => Visual Studio 2019
- Problem classification => PQT
##### Detailed desc…
-
### 🐛 Describe the bug
This is a toy program that uses a queue to get the tensors generated by its subprocess.
```python
import torch
from torch.multiprocessing import Queue, Process
def simple…
-
My system specification :
PyTorch :- 1.11.0,
Python :- 3.8.8,
Cudatoolkit :- 11.3.1,
CUDA :- 11.2,
cudnn :- 8.2.1
I am trying to install torch-points3d, but unfortunately, I am getting …