omnimvs_pytorch icon indicating copy to clipboard operation
omnimvs_pytorch copied to clipboard

Published 20 hours ago • verified publisher icon matsuren

Loss value getting saturated after one epoch

Open Uma1699 opened this issue 2 years ago • 2 comments

The loss value gets saturated to 0.0010 after one epoch and does not change after that. How can I resolve this issue? Please help

Uma1699 avatar Apr 01 '22 07:04 Uma1699

Hello, has your problem been solved later? I also encountered the same situation!

Chen-z-s avatar Jun 08 '22 01:06 Chen-z-s

Hello, has your problem been solved later? I also encountered the same situation!

Because the currently downloaded label data (.tiff format) is an inverse depth map, there is no need to convert it. And the load_invdepth() function of the code calculates this conversion process.

zeki096 avatar Apr 12 '23 16:04 zeki096