About the cuda error.

[WangZhouTao]

Hello, thank you for your good work. when I try to train the network with S3DIS, I get this error report: tensorflow.python.framework.errors_impl.NotFoundError: /home/logic/Desktop/2021Project/BAAF-Net/utils/sampling/tf_sampling_so.so: undefined symbol: __cudaPushCallConfiguration Do you have any suggestions？

[ShiQiu0419]

could you please specify your cuda location by changing the line: https://github.com/ShiQiu0419/BAAF-Net/blob/663d1681d4d05ad3caaacd98e6dedfdc9caa4930/utils/sampling/compile_ops.sh#L6

then recompile /utils/sampling/compile_ops.sh

[M-leng]

I compile it with Cuda 10.1. I get this error report: tensorflow.python.framework.errors_impl.NotFoundError: /home/lzy/BAAF-Net-main/utils/sampling/tf_sampling_so.so: undefined symbol: cudaLaunchKernel Is Cuda 10.0 only, 10.1 not?

[LiXinZhana]

Hello, thank you for your good work. when I try to train the network with S3DIS, I get this error report: tensorflow.python.framework.errors_impl.NotFoundError: /home/logic/Desktop/2021Project/BAAF-Net/utils/sampling/tf_sampling_so.so: undefined symbol: __cudaPushCallConfiguration Do you have any suggestions？

Hello! I have the same question. Have you sovled the problem?

[LiXinZhana]

could you please specify your cuda location by changing the line:

https://github.com/ShiQiu0419/BAAF-Net/blob/663d1681d4d05ad3caaacd98e6dedfdc9caa4930/utils/sampling/compile_ops.sh#L6

then recompile /utils/sampling/compile_ops.sh

I specify my cuda location and recompile, but still have this problem. could you give me some advice?

[ShiQiu0419]

could you please specify your cuda location by changing the line: https://github.com/ShiQiu0419/BAAF-Net/blob/663d1681d4d05ad3caaacd98e6dedfdc9caa4930/utils/sampling/compile_ops.sh#L6

then recompile /utils/sampling/compile_ops.sh

I specify my cuda location and recompile, but still have this problem. could you give me some advice?

would you please show me your versions of tensorflow and cudatoolkit in your environment by running conda list?

[LiXinZhana]

could you please specify your cuda location by changing the line: https://github.com/ShiQiu0419/BAAF-Net/blob/663d1681d4d05ad3caaacd98e6dedfdc9caa4930/utils/sampling/compile_ops.sh#L6

then recompile /utils/sampling/compile_ops.sh

I specify my cuda location and recompile, but still have this problem. could you give me some advice?

would you please show me your versions of tensorflow and cudatoolkit in your environment by running conda list? I use pip and my version of tensorflow is 1.13.1

[ShiQiu0419]

could you please specify your cuda location by changing the line: https://github.com/ShiQiu0419/BAAF-Net/blob/663d1681d4d05ad3caaacd98e6dedfdc9caa4930/utils/sampling/compile_ops.sh#L6

then recompile /utils/sampling/compile_ops.sh

I specify my cuda location and recompile, but still have this problem. could you give me some advice?

would you please show me your versions of tensorflow and cudatoolkit in your environment by running conda list? I use pip and my version of tensorflow is 1.13.1

You may create a new environment and install the tensorflow by running: conda install tensorflow-gpu==1.13.1 then make sure that the installed cudatoolkit version is consistent with the cuda on your device.

For example, on my PC: tensorflow-gpu 1.13.1 cudatoolkit 10.0.130 CUDA_ROOT=/usr/local/cuda-10.0