diffindscene
diffindscene copied to clipboard
torchsparse running issue
Hi, tanks for your awesome work. I follow your installation guid using - pytorch 1.13.1 - torchsparse 1.4 - cuda 11.7
but when i running the infernce code , i get error
AttributeError: module 'torchsparse.backend' has no attribute 'build_kernel_map_subm_hashmap
do you have another setting? my machine is 4090 cuda with driver 12.4