Closed drakgoku closed 1 year ago
They provide a readme in colab, I'm not sure if it is helpful to you.
Windows platform is not supported very well. My suggestion is that Linux is the first choice to be considered.
@drakgoku Windows WSL2 support CUDA and docker.You can try to build a wsl2 environment and use the docker install function.
Thanks @Dandelionym & @Tom0072 for the answers!!
Yes, we only have the model tested on Linux. Deploying this code on Windows is out of the scope of research and would involve a lot of engineering effort.
Nevertheless, we do plan to release it in the NVIDIA Omniverse soon and it should also support Windows in the future! check this blog for more information https://developer.nvidia.com/blog/rapidly-generate-3d-assets-for-virtual-worlds-with-generative-ai/
Thanks @SteveJunGao, hope it will be released soon!
Was progress made?
I think it is necessary to reopen the thread. Can someone reopen the thread?
This is research code, not an app. It's possible to run it on Windows using WSL, but it's not really something that would ever have a simple interface.
There are many things I don't understand. I have read the ReadMe. The problem is that in an S.O. Windows I don't know how to do it. Surely there is some way.
I have read: https://github.com/nv-tlabs/GET3D/issues/76
Is it so difficult that instead of having to put commands, you create a simple interface that is executable for multiple platforms?
Can you imagine playing Mario Bross with commands?