how many memory does the Nest Unet require when dealing with 3D image?

bigmb / Unet-Segmentation-Pytorch-Nest-of-Unets

Implementation of different kinds of Unet Models for Image Segmentation - Unet , RCNN-Unet, Attention Unet, RCNN-Attention Unet, Nested Unet

MIT License

1.87k stars 345 forks source link

how many memory does the Nest Unet require when dealing with 3D image? #27

Closed Kay-Shaw closed 4 years ago

Kay-Shaw commented 4 years ago

I modified the network to deal with 3D medical image with a GPU(12G), but it feeds back "CUDA out of memory. Tried to allocate 1.50 GiB (GPU 0; 11.91 GiB total capacity; 9.84 GiB already allocated; 1.45 GiB free; 21.18 MiB cached)"

so how many GPU memory does the Nest Unet require when dealing with 3D image?

bigmb commented 4 years ago

Nest Unet requires a lot of memory. (But less than R-Attention-CNN Unet) Whats the size of your input? Also are you using patches or complete 3D image as an input?

Kay-Shaw commented 4 years ago

i used resized 128^3 image as input, maybe I should try smaller size, but that would lead to poor performance

bigmb commented 4 years ago

Try it with a smaller size and check if the nested unet works. If it works , then the current issue is due to allocation of memory.