coldgemini
coldgemini
I got the same issue, the loss suddenlly goes to NAN at some iter. Did you guys solved this problem by fixing this PAF bug?
isn't this exactly why ndc coord range should be [0,1] ? since forward conversion is world coord to ndc [0,1] then backward conversion should be ndc[0,1] to world coord?
how did your get the CV papers on that day? by title?