search

drethage / fully-convolutional-point-network

Fully-Convolutional Point Networks for Large-Scale Point Clouds
MIT License
85 stars 22 forks source link

Insufficient memory #5

Closed hanxuel closed 5 years ago

hanxuel commented 5 years ago

Hi, whenever I do data.prepare to extract samples from scene, it gives error that memory is insufficient. When I reduce the number of scenes in the train_split set, it starts to work. I'd like to know how much RAM do you use for training? I use 1 core with memory 14000M, but it seems far less sufficient.

drethage commented 5 years ago

Hi hanxuel, I worked on a machine with 8 cores and 32GB of RAM. It may be necessary to have a bit more RAM. More cores will certainly speed up dataset preparation too.