shunsuke yokokawa
shunsuke yokokawa
just you don't have enough GPU memory. use bigger GPU or use small batch size 1
in main.py KITTI dataset has minus disparity or zero disparity which is impossible mask = (disp_true < args.maxdisp) mask = (disp_true > 0) #you need to have this line!!!!!!! there...
Thanks for fixing it
@hhaAndroid batch data is a custom type instead of torch Tensor, that's why it doesn't get acceleration? https://pytorch.org/docs/stable/data.html#memory-pinning 