＋１Cannot install the dependencies for the virtual environment about rignet HOT 3 CLOSED

Hi. I tried pytorch 1.5+ pytorch_gemetric1.5 on my machine and it works, so I think you can try install latest pytorch and pytorch_geometric (v 1.6.0). Since you system cuda is 9.1, I think it is better to use cudatoolkit in conda. Did you do exactly this?

conda install pytorch torchvision cudatoolkit=10.2 -c pytorch

After this command could you try:

python -c "import torch; print(torch.__version__)"
python -c "import torch; print(torch.version.cuda)"

What's the outputs?

Thanks!
I have solved this problem:First,I change another GPU server with CUDA 10.0

• conda install pytorch==1.4.0 torchvision==0.5.0 cudatoolkit=10.0 -c pytorch
• pip install torch-scatter==latest+cu100-f https://pytorch-geometric.com/whl/torch-1.4.0.html
• pip install torch-sparse==latest+cu100 -f https://pytorch-geometric.com/whl/torch-1.4.0.html
• pip install torch-cluster==latest+cu100 -f https://pytorch-geometric.com/whl/torch-1.4.0.html
• pip install torch-spline-conv==latest+cu100 -f https://pytorch-geometric.com/whl/torch-1.4.0.html
• pip install torch-geometric

Thanks for your help~

from rignet.

Why I can successfully install torch1.3.0+cu100 and use it, but unsuccessfully install torch-scatter、torch-sparse、torch-cluster??

from rignet.

Hi. I tried pytorch 1.5+ pytorch_gemetric1.5 on my machine and it works, so I think you can try install latest pytorch and pytorch_geometric (v 1.6.0). Since you system cuda is 9.1, I think it is better to use cudatoolkit in conda. Did you do exactly this?

conda install pytorch torchvision cudatoolkit=10.2 -c pytorch

After this command could you try:

python -c "import torch; print(torch.__version__)"
python -c "import torch; print(torch.version.cuda)"

What's the outputs?

from rignet.