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

Complie PyTorch Bindings with float32 precision. #51

Closed
srxdev0619 opened this issue Feb 27, 2022 · 2 comments
Closed

Complie PyTorch Bindings with float32 precision. #51

srxdev0619 opened this issue Feb 27, 2022 · 2 comments

Comments

@srxdev0619
Copy link

Hi,

First up, thanks for this is great work! I was just wondering if there's a way to compile tiny-cuda-nn and its pytorch bindings to use float32 by default? Thanks!

@srxdev0619 srxdev0619 changed the title Complie PyTorch with float32 precision. Complie PyTorch Bindings with float32 precision. Feb 27, 2022
@Tom94
Copy link
Collaborator

Tom94 commented Feb 28, 2022

Hi there, you can go into include/tiny-cuda-nn/common.h and change

#define TCNN_HALF_PRECISION (!(TCNN_MIN_GPU_ARCH == 61 || TCNN_MIN_GPU_ARCH <= 52))

to

#define TCNN_HALF_PRECISION 0

Cheers!

@Tom94 Tom94 closed this as completed Feb 28, 2022
@srxdev0619
Copy link
Author

Worked perfectly, 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

2 participants