You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When running cell_inference.py on larger WSI files (currently just piloting this on individual slides), the postprocessing stage does not complete. I can also see that memory usage increases from ~20 gb max during inference to the maximum amount that I've requested from my HPC after inference has finished (most recently went up to 63 gb):
No errors are thrown, but the session times out at this stage. Do you have any advice for resolving this issue? I'm able to run inference on smaller files without any problems. I'm running the PanNuke-pretrained CellViT-SAM-H-x40.pth model with --geojson.
The text was updated successfully, but these errors were encountered:
Sorry, I think the postprocessing in the current stage is very memory intensive. Our machines for inference have large ram. Maybe you need to change the postprocessor. I have this on my list for a new version, but have not find the time to implement it on my own
When running cell_inference.py on larger WSI files (currently just piloting this on individual slides), the postprocessing stage does not complete. I can also see that memory usage increases from ~20 gb max during inference to the maximum amount that I've requested from my HPC after inference has finished (most recently went up to 63 gb):
2024-12-07 19:08:55,566 [INFO] - Detected cells before cleaning: 913943
2024-12-07 19:08:55,566 [INFO] - Initializing Cell-Postprocessor
No errors are thrown, but the session times out at this stage. Do you have any advice for resolving this issue? I'm able to run inference on smaller files without any problems. I'm running the PanNuke-pretrained CellViT-SAM-H-x40.pth model with --geojson.
The text was updated successfully, but these errors were encountered: