-
Notifications
You must be signed in to change notification settings - Fork 585
Description
I am using cellpose GUI using python 3.8. I tried to use 3D functionality but I get the following memory error:
(BIGMEOW) C:\Users\meow>cellpose
2021-08-26 13:44:43,566 [INFO] WRITING LOG OUTPUT TO C:\Users\meow.cellpose\run.log
2021-08-26 13:44:47,822 [INFO] ** TORCH CUDA version installed and working. **
X:/A/Ed/output/cellimage.tif
138 (411, 843, 3)
removed all cells
cyto
2021-08-26 13:45:30,836 [INFO] ** TORCH CUDA version installed and working. **
2021-08-26 13:45:30,837 [INFO] >>>> using GPU
estimated diameter of cells using cyto model = 41.7 pixels
removed all cells
cyto
2021-08-26 13:45:47,229 [INFO] ** TORCH CUDA version installed and working. **
2021-08-26 13:45:47,230 [INFO] >>>> using GPU
using model cyto
2021-08-26 13:45:47,346 [INFO] ~~~ FINDING MASKS ~~~
2021-08-26 13:45:47,346 [INFO] multi-stack tiff read in as having 138 planes 3 channels
2021-08-26 13:45:49,362 [INFO] running YX: 138 planes of size (411, 843)
2021-08-26 13:45:49,442 [INFO] 0%| | 0/35 [00:00<?, ?it/s]
2021-08-26 13:45:50,568 [INFO] 0%| | 0/35 [00:01<?, ?it/s]
NET ERROR: CUDA out of memory. Tried to allocate 196.00 MiB (GPU 0; 8.00 GiB total capacity; 5.69 GiB already allocated; 0 bytes free; 6.30 GiB reserved in total by PyTorch)
I have two GPUs:
GPU 0 - Intel UHD Graphics 630
GPU 1 - Nvidia GeForce RTX 2070 with Max-Q Design (8 GB)
May be it is using GPU 0 by default, if that is the case how to change it so that it uses GPU 1. The size of the image I am using is also not that big (only 46 MBs)
Do you have any suggestions?