otakudj

Same question

I think changing source code is a dangerous method. I try to change gcc & g++ version to 9.x, and successfully build the code, although there are still many warnings....

Do you guys resolve? I have the same problem with 3090 @chl2 @jaharkes 

it's the [error](https://forums.developer.nvidia.com/t/cuda-11-3-tex2dlod-causes-an-internal-compiler-error-with-std-c-17/175531/5?u=ydj1130) of cu113. The solution is change cuda to another version (e.g. 11.7 or 11.1), and change the corresponding libtorch, then you will succeed. Good luck @sumanttyagi @ngrxmu

I have the same problem. When I use single 3090 gpu running the demo lego, it takes ~9secs an epoch.  Howerver, when I use 3090x4 to train the model,...

I have the same error, you can try to change 「sample step」 to ~50, then the error disappear but the result is not so good. @CultureAddiction

> > We are actively preparing the code of reconstruction model and the codes will be released in two weeks. > > Hi, when will the training code be released...

Wow! Looking forward for your powerful model! 👍🏻And thank you for your reply, i wish you could @ me when the texture part ready.