Julien
Julien
I can confirm it works with Qt5 projects, but it fails to find most of Qt4 included headers
I can't see any improvements after updating the powershell scripts
I would say: ignore it and print some warning message
can you give a look at this commit? https://github.com/pauloromeira/onegram/compare/master...neFAST:34e7f348f583c99abb74308d554c7ade4a9f6b1d#diff-97323fbaa4abdcae6b08f8df4c581b41
@bygreencn Interesting point of view, may I ask if you have any network/papers to recommend when it comes to detailed defect segmentation (ie. semantic segmantation, not classification)?
> [Render](https://render.com/) worked well for me. May I ask more details on how you deployed it to render? Thanks in advance
What about submitting a PR?
FYI this PR never went through: https://github.com/qubvel/segmentation_models.pytorch/pull/474
Just get rid of the activation layer?