train question

[WS517]

Thanks for your great work. I tried to train with my own data, the data I dropped into the network is in the format of xyzrgb semantic labels as well as instance labels, but while training the following problem occurs even though I can loop: RuntimeError: cuda runtime error (9) : invalid configuration argument at /pytorch/aten/src/THC/THCTensorMathCompare.cuh:82
How can I solve this problem?