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

bfloat16 support for stable half precision training #7

Open
KickCellarDoor opened this issue Feb 8, 2024 · 0 comments
Open

bfloat16 support for stable half precision training #7

KickCellarDoor opened this issue Feb 8, 2024 · 0 comments

Comments

@KickCellarDoor
Copy link

Thanks for the awesome job! Currently nr3d_lib supports float64, float32 and float16, I think bfloat16 support can be introduced without too much effort and for my experience it can stabilize NeRF training with half precision. Hope you can consider this option later, thanks!

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

1 participant