Tuatini Godard

Thanks, but even considering this I tried few options which didn't seem to help :( Like: ``` let pyshell = new PythonShell("-m", { mode: 'text', scriptPath: path.join(src_path, "src"), args: [`src.main`,...

I think I got it: ``` let pyshell = new PythonShell(path.join("src", "main.py"), { mode: 'text', scriptPath: src_path, args: [`eval`, `--input_wsi=${currentWSI.path}`, `--output=${dir}`, `--pred_mode=${pred_mode}`, `--downsample=${downsample}`, `--model=${model}`, `--stdlog`] }); ``` Setting the `scriptPath`...

Here is another solution that is a bit hacky but works to clean up the working directory before attempting any checkout: ``` jobs: cleanup: runs-on: self-hosted container: image: ubuntu:latest steps:...

Hi @jcupitt , Thanks a lot for this response! We did some pretty data augmentation in fact during training, but during eval we still end up with less good predictions...

Turns out the `kernel="linear"` was in fact the culprit. It changes everything for me now that you pointed me in the right direction. My friend and I spent countless nights...

I ended up doing: `ln -s /home/ekami/Programs/anaconda3/envs/dl/lib/libMagick++-7.Q16HDRI.so /home/ekami/Programs/anaconda3/envs/dl/lib/libMagickCore-7.Q16HDRI.so.6` And it kinda worked, not sure if that's a sustainable solution tho.

@sebastian-luna-valero I'd say `imagemagick` since that's the one missing from the dynamic dependencies

@sebastian-luna-valero It works indeed!

Hello, Please also ensure the conda package uses the `pixman=0.36.0` dependency and not the 0.38.0, the reason is because the version 0.38.0 create corrupted images with various levels of zoom....

Thank you so much @jcupitt !! I have another question: If I use `compression='jpeg', Q=85` wouldn't I loose on image quality on top of the jpeg2000 compression the SVS files...