Official PyTorch implementation of Superpoint Transformer introduced in [ICCV'23] "Efficient 3D Semantic Segmentation with Superpoint Transformer" and SuperCluster introduced in [3DV'24 Oral] "Scalable 3D Panoptic Segmentation As Superpoint Graph Clustering"
MIT License
560
stars
72
forks
source link
Proposition to help debugging the SPT installation for non supported CUDA versions #156
Superpoint Transformers rely on ( a lot ) of highly GPU optimized dependency ( ex: Pytorch and Pytorch geometric) . Those dependency installlation can be quite tricky if your CUDA versions differ from the one provided in the readme.md.
I mainly met 4 errors when I tried to install SPT on some device:
torch scatter wasn't installed with GPU support
FRNN failed in installation ( mostly because of pytorch issue)
pgeof isn't installed
I had issue with data.keys() or data.keys depending on the version of pytorch geometrics.
I updated the readme and a created a folder " setup". The install script has been moved to this folder and I proposed a script " check_dependency.sh" that try to identify each one of the above errors .
Those tests is not exhaustive but it helped me tremendously to identify when ( and where) install.sh failed when I had to change clusters.
Before submitting
[x ] Did you make sure title is self-explanatory and the description concisely explains the PR?
[x ] Did you make sure your PR does only one thing, instead of bundling different changes together?
[ x] Did you list all the breaking changes introduced by this pull request?
[x ] Did you test your PR locally with pytest command?
[ ] Did you run pre-commit hooks with pre-commit run -a command?
What does this PR do?
Superpoint Transformers rely on ( a lot ) of highly GPU optimized dependency ( ex: Pytorch and Pytorch geometric) . Those dependency installlation can be quite tricky if your CUDA versions differ from the one provided in the readme.md.
I mainly met 4 errors when I tried to install SPT on some device:
Those tests is not exhaustive but it helped me tremendously to identify when ( and where) install.sh failed when I had to change clusters.
Before submitting
pytestcommand?pre-commit run -acommand?Did you have fun?
Make sure you had fun coding 🙃