André Pedersen

Definitely. It should just be an argument to set in Model.fit() or similar. GA is already available in [pytorch-lightning](https://pytorch-lightning.readthedocs.io/en/stable/advanced/training_tricks.html), and others have added support for GA in their framework. About...

> I was able to fix this by adding in casts to the appropriate dtype for the the `build` and `__call__` functions Interesting, @zibbini. LGTM. Can run some checks tomorrow.

@zibbini Just ran some experiments, but ran into more issues. See [gist](https://colab.research.google.com/gist/andreped/7e8bf14d7db851bdf4f43d35f6c9c787/untitled503.ipynb#scrollTo=izDg48QF1YDI). After adding your modifications, it seems to work, but it fails to load the pretrained weights.

Suggestion (1) is something I have also wanted for a long time. Same with smoother movements including deaccelerations, similar to what QuPath (and OpenSeadragon used for LearnPathology) is doing.

> Thanks for reporting, the issue is that we forgot to upload the file here: Has this now been fixed for any of the new releases? --- EDIT: It does...

@ajr82 this fix has now been made available in the latest release: https://github.com/AICAN-Research/FAST-Pathology/releases/tag/v1.1.0 All segmentation results should now be stored in the same pyramidal TIFF format, which in turn means...

Just tested again with the latest release of FP on Win11, and I no longer see this issue. I tried both visualizing and running nuclei segmentation. Might have been a...

I have made a draft for Bison in PR https://github.com/vanna-ai/vanna/pull/264. But I guess this is more about improving documentation to make it easier for others to add support for a...

I would think this would fail: https://github.com/girder/large_image/blob/master/large_image/tilesource/base.py#L754

Just mentioning that I am now able to run inference with a deep learning model using my own custom plugin (https://github.com/andreped/FP-dsa-plugin), but the end users mostly have their images in...