Open hoangcuongbk80 opened 5 years ago
I got a same error,and hava the same GPU, is there anyone know how to deal it?
I made it work, but do not remember how exactly I did it. You can try what I fixed from here: https://github.com/hoangcuongbk80/Object-RPE/tree/master/DenseFusion
I made it work, but do not remember how exactly I did it. You can try what I fixed from here: https://github.com/hoangcuongbk80/Object-RPE/tree/master/DenseFusion
thank you very much,it's worked for me , although came new problem but I solved it
Great! It might come with: Error: AttributeError: module 'lib.knn.knn_pytorch' has no attribute 'knn' Solution: rebuild and unzip egg file then copy *so file and knn_pytorch.py to /densefusion/lib/knn
Great! It might come with: Error: AttributeError: module 'lib.knn.knn_pytorch' has no attribute 'knn' Solution: rebuild and unzip egg file then copy *so file and knn_pytorch.py to /densefusion/lib/knn
yes, I met this problem and solved it by this way
@hoangcuongbk80 could you tell me how to rebuild that? I mean what the command is? Thanks
@704051419 You can find the solution in the file below: https://github.com/hoangcuongbk80/Object-RPE/blob/master/DenseFusion/note.txt
@MrLuer @hoangcuongbk80 @704051419 @yukezhu Hi, after rebuilding that, it caused a new problem, do you know how to deal with it? RuntimeError: aborting at /home/llcrobot/Robby/PoseEstimation/2019_CVPR_DenseFusion/DenseFusion-Pytorch-1.0/lib/knn/src/knn.h:45 Segmentation fault (core dumped)
Did you copy *so file and knn_pytorch.py to /densefusion/lib/knn?
Yes,I have copied it
I made it work, but do not remember how exactly I did it. You can try what I fixed from here: https://github.com/hoangcuongbk80/Object-RPE/tree/master/DenseFusion
hi, I have the same problem as you. My system also same as you. But as I follow you suggestion , I still get some error: ImportError: torch.utils.ffi is deprecated. Please use cpp extensions instead.
I know this is the Pytorch version problem, but I follow the suggest and my pytorch version is torch==1.0.1.post2 . I can't solve the problem. I hope you can help me , thanks!
@RobbyHuang @SCMStone You might try the docker image I built here. It works for me.
@RobbyHuang @SCMStone You might try the docker image I built here. It works for me.
thanks, I will try it!
@hoangcuongbk80 @MrLuer
I got the same issue:
pred = torch.add(torch.bmm(model_points, base), points + pred_t) RuntimeError: cublas runtime error : the GPU program failed to execute at /pytorch/aten/src/THC/THCBlas.cu:441
But I don't know how to solver it by https://github.com/hoangcuongbk80/Object-RPE/tree/master/DenseFusion Could you elaborate it in details?
Thank you!
I used Pytorch-1.0 branch and ran ./experiments/scripts/train_ycb.sh. Then I got the error:
pred = torch.add(torch.bmm(model_points, base), points + pred_t) RuntimeError: cublas runtime error : the GPU program failed to execute at /pytorch/aten/src/THC/THCBlas.cu:441
My system: