Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Computation Requirements #34

Open
rohit7044 opened this issue May 5, 2024 · 3 comments
Open

Computation Requirements #34

rohit7044 opened this issue May 5, 2024 · 3 comments

Comments

@rohit7044
Copy link

Wonderful project guys!! Kudos to the team!!
While running the code in windows, I got an CUDA OOM error.
I am using 6GB VRAM. Is it not enough?

@maturk
Copy link
Owner

maturk commented May 5, 2024

Hi @rohit7044, which scene are you testing? I reccommend trying Replica dataset; since for you the main bottleneck is scene size (number of gaussians required to represent the scene) and cached images size. I cache 64 images for Replica scenes, but some scenes e.g from mushroom can have over 600 imgs.

@rohit7044
Copy link
Author

I am using MipNeRF 360 dataset Room scene and it has more than 1000+ images in each folder.
FYI Splatfacto works perfectly on this scene.

@maturk
Copy link
Owner

maturk commented May 5, 2024

@rohit7044 dn-splatter needs a bit more ram than splatfacto due to storing and rendering normal maps. I can test the mip-nerf scene once I have time, and report how much ram it consumes. In the meantime, you could try a smaller scene for testing.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants