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

Running Demo File Shared Libraries issue #21

Open
TamourBilal6 opened this issue Jun 21, 2021 · 4 comments
Open

Running Demo File Shared Libraries issue #21

TamourBilal6 opened this issue Jun 21, 2021 · 4 comments

Comments

@TamourBilal6
Copy link

I am facing this shared libraries error when i am trying to start demo.py

If you depend on functionality not listed there, please file an issue.

Traceback (most recent call last):
File "demo.py", line 12, in
from modules.models_mvp2m import MeshNetMVP2M as MVP2MNet
File "/content/gdrive/MyDrive/new/Pixel2MeshPlusPlus/modules/models_mvp2m.py", line 9, in
from modules.losses import mesh_loss, laplace_loss
File "/content/gdrive/MyDrive/new/Pixel2MeshPlusPlus/modules/losses.py", line 6, in
from modules.chamfer import nn_distance
File "/content/gdrive/MyDrive/new/Pixel2MeshPlusPlus/modules/chamfer.py", line 4, in
nn_distance_module = tf.load_op_library('./external/tf_nndistance_so.so')
File "/tensorflow-1.15.2/python3.7/tensorflow_core/python/framework/load_library.py", line 61, in load_op_library
lib_handle = py_tf.TF_LoadLibrary(library_filename)
tensorflow.python.framework.errors_impl.NotFoundError: ./external/tf_nndistance_so.so: cannot open shared object file: No such file or directory

@topinfrassi01
Copy link
Contributor

From where do you run the python script? Because tf_nndistance_so.so is loaded from a relative path. You should run the script directly from the project's root.

If this isn't the problem, then do these files exist? If not, you should look at your cuda version (you can see it using nvidia-smi inside the terminal). If you use cuda9.0 you can copy the files from the archive-cuda-9 folder, if you run 10.0, then copy from the 10.0 folder.

If you run any other version, you'll need to compile the cuda files using the procedure described inside the Makefile file in the external folder.

If this still doesn't work, you can reply to this thread I'll try to help you out.

@TamourBilal6
Copy link
Author

I have made changes in the makefile and given path to the CUDA and Nvidia files but still facing the same issue

@TamourBilal6
Copy link
Author

Hi @topinfrassi01 I have sort out the issue by copying files from tf_nndistance_so.so from CUDA 9 archieve to external folder however at the very next i found out "libcudart.so.9.0: cannot open shared object file: No such file or directory" this issue. Can u plz help me out to resolve this issue. Regards

@topinfrassi01
Copy link
Contributor

If libcudart 9.0 can't be found, I believe it is because you are not using CUDA 9.0. Could you maybe give me the given path for CUDA (that you put inside the Makefile) and the output of the nvidia-smi command?

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

2 participants