Question about cuda version in neucon

[Cli98]

Hi @achao2013 ,

I know that it is rude to ask question across repo. However I run into issues in neuralrecon called "no valid points: scale 0 when runing example data" #40 . I see that you solved this issue. Can you briefly describe how you changed cuda to solve the issue?

Thank you if you can help me.

[achao2013]

i remember it's caused by torchsparse, its version should be 1.4 and cuda version should be 10.2, and you can look up the versions of other dependent libraries in my environment documents of this repository