search

krrish94 / nerf-pytorch

A PyTorch re-implementation of Neural Radiance Fields
Other
872 stars 123 forks source link

Getting 'Killed' whenever trying to train #15

Closed joy-alen closed 3 years ago

joy-alen commented 3 years ago

Hi, Whenever I try to train I am getting 'Killed' as an output. This is the same even when trying to cache rays from the output. Can I have some suggestions for resolving this problem? Thanks in advance.

krrish94 commented 3 years ago

This might usually mean that python is using more than intended resources (memory usually). If you're running this on a server, be sure to request enough memory

joy-alen commented 3 years ago

I'm currently running it on my system. What should be the required amount of memory?

krrish94 commented 3 years ago

In general that's hard to answer. For tiny configs (eg. Blender datasets, 100 x 100 image resolution, 32 or so samples per ray), I think a computer with 10-12 GB memory, with a GPU of around 2-4 GB memory should be easily able to run this

joy-alen commented 3 years ago

Thank you for the answer.