skhadem / 3D-BoundingBox

PyTorch implementation for 3D Bounding Box Estimation Using Deep Learning and Geometry
MIT License
435 stars 96 forks source link

testing on cpu added #20

Open KunikaValecha opened 3 years ago

KunikaValecha commented 3 years ago

This code can can run for small number of images on cpu. It has been tested successfully on pytorch version = 1.4.0+cpu

skhadem commented 3 years ago

Hi! Sorry for the delay in responding to this. I see that the only thing that has changed is removing the calls to .cuda(). What do you think about instead adding a flag (something like --cpu_only) directly to Run.py? Then, whenever there is a call to cuda(), if this new flag is set we can keep the variables on the spu. This way we avoid duplicated code. Let me know if you have any questions!

KunikaValecha commented 3 years ago

Yes, that would work.

skhadem commented 3 years ago

Do you want to make the changes in your branch and push directly to this PR?