Pietro Antonio Cicalese
Pietro Antonio Cicalese
How many workers per gpu? Batch size? Also check if you are computing gradients (i.e. make sure you are in eval mode).
Pytorch uses some memory for meta-data and CUDA context management, and CUDA code itself also uses a significant amount of cpu memory (has to do with unified memory I think)....
@Thevakumar-Luheerathan please check your config files, you will see a `num_workers` key in your dataset dicts.
Yes - it is quite robust to source scanners. I have not tested it with all scanner data structures, but it appears to be much more reliable than openslide. Additionally,...
Yes I did. You can load it in parts or use ROIs to load a section only. I can share a piece of code I used to do that later...
Created new conda env here with `pytorch-lightning==1.6.1` and it seems to work, after running `conda env remove -n ldm` and recreating the env with new .yml as pointed out by...
@leodvd were you able to resolve this? Working on a similar problem.
Thanks so much @leodvd ! So if I am understanding correctly, you project the colorspace onto the mask; did you have a specific approach to accomplish this? How did you...
@leodvd my current solution is to encode the mask using the 4 channel mean, and then decode the generated mask later using the generated rgb pixel values. Thanks a lot...
You need to set the n_classes input in your cond_stage_config. See your ldm.modules.encoders.modules.ClassEmbedder if you are using that. They use the default 1000 because they only used that module on...