talk2bev icon indicating copy to clipboard operation
talk2bev copied to clipboard

CUDA out of memory

Open weizhou1991 opened this issue 4 months ago • 0 comments

Hi there,
When I was using your "generate_base.py" script file, I got "CUDA out of memory" when exex line 211 "output = model( image.to(device), intrinsics.to(device), extrinsics.to(device), future_egomotion.to(device) )", I am using Tesla V100, 16 GB. I am wondering which type you was used. How much GPU memory you used for that script ?

Best,

weizhou1991 avatar Mar 14 '24 08:03 weizhou1991