search

POSTECH-CVLab / point-transformer

This is an unofficial implementation of the Point Transformer paper.
508 stars 99 forks source link

NaN Loss #39

Open matzel9 opened 2 years ago

matzel9 commented 2 years ago

Hi, first of all thank you for the great repository and the efficient implementation. Does someone else encounter a NaN loss due to the sum operation in the TransitionUP module of the decoder 5. I removed the two linear layers and the training works fine. Please comment if you solved the issue.

vardeep-sandhu commented 2 years ago

I am facing the same issue as well. Did you figure out how to solve this error?

winnechan commented 2 years ago

@vardeep-sandhu Did you find out the reason? Thanks