out put is broken at png with alpha channel

[taizan]

the model is trained for grayscale img
so we have to convert or alert for png with alpha channel img

[Noclue12]

I have one question. Not sure where to ask it so I try here and hope I get a answer. If I install this localy. Will paintsschainer be able to output larger images? Like if my base image is 3000 to 3000 pixels would paintschainer put out a colored version of it with the same size / resolution?

I tried the docker installation route but that gave me only the web version it seems!?

also about installing this on windows. Step 3 is confusing to me. “Deep learning SDK” What exactly do I need to download and install here since there are a lot of tools under this tab im kind of confused.
Hope you can help me. thanks

[taizan]

Dont ask entirely different topic from title. Make new Isuues.

• the input image size is decided in the code at img2imgdataset.py. so if you want to change you have to change the code.
• you need CUDA & cuDNN