Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

CUDA 10.1 #16

Closed
gabriel-lucas opened this issue Aug 17, 2022 · 4 comments
Closed

CUDA 10.1 #16

gabriel-lucas opened this issue Aug 17, 2022 · 4 comments

Comments

@gabriel-lucas
Copy link

gabriel-lucas commented Aug 17, 2022

Hi,

Thanks for the nice work you share, I find it very interesting.

I´m trying to install it but I found that I´m running a different cuda version 10.1.

nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2019 NVIDIA Corporation
Built on Sun_Jul_28_19:07:16_PDT_2019
Cuda compilation tools, release 10.1, V10.1.243

The environment.yml in the repository installs newer version of cudatoolkit (11.3.1=h2bc3f7f_2) and this ends up in a pytorch3d error when running demo.py.

from pytorch3d import _C
ImportError: libtorch_cuda_cu.so: cannot open shared object file: No such file or directory

It looks related with this.
facebookresearch/pytorch3d#1013 (comment)

Some packages versions should be changed to adapt to the cuda version.

Could someone point me in the right direction?

Best regards.

@lzqfdd
Copy link

lzqfdd commented Aug 17, 2022

Hello, are you also running this code? When you encounter that problem, you can try to create an environment first, and then install CUDA according to your server to see if it is compatible. The premise is that this code supports older CUDA versions. I have a problem that I have downloaded antelopev2, but I still go back to the website to download it when running. Have you met? If you encounter the same error, can you help me solve it? Thank you.

@Zielon
Copy link
Owner

Zielon commented Aug 17, 2022

The env file is using a nightly version of pytorch3d because I had problems with my RTX3090. One of the solutions would be to remove pytorch3d-nightly from dependency and install if from the pytroch3d channel also adapting the cuda toolkit version.

@gabriel-lucas
Copy link
Author

Thanks @Zielon, you were of great help :)

I tested your proposal and it worked flawlessly.

@gabriel-lucas
Copy link
Author

I think it´s better if you open a new issue describing problem you are facing in detail.

Hello, are you also running this code? When you encounter that problem, you can try to create an environment first, and then install CUDA according to your server to see if it is compatible. The premise is that this code supports older CUDA versions. I have a problem that I have downloaded antelopev2, but I still go back to the website to download it when running. Have you met? If you encounter the same error, can you help me solve it? Thank you.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants