PyCDA icon indicating copy to clipboard operation
PyCDA copied to clipboard

Published 20 hours ago • verified publisher icon lianqing11

aux_loss

Open Xcrid opened this issue 4 years ago • 2 comments

Hi,

You use two losses for your training, yet all your decoders return the main output only, the aux one being None.

Which one should I add ?

Xcrid avatar Apr 24 '20 17:04 Xcrid

Maybe I misunderstood, but doesn't _ discard the variable ?

Xcrid avatar Apr 25 '20 22:04 Xcrid

Hi @Xcrid , sorry for the late reply. The aux loss is from the PSPNet. the implementation of PSPNet is from https://github.com/CSAILVision/semantic-segmentation-pytorch. Maybe It is confused for the variable name. I will modify it once I have time.

lianqing11 avatar May 01 '20 09:05 lianqing11