Update Montgomery dataset

Updated the Montgomery dataset and associated tests to fit the new data handling design implemented in the add-datamodule branch

assigned to @dcarron

changed title from Update montgomery to Update Montgomery dataset

added 2 commits

• bb035416 - [data.montgomery.loader] Use load_pil() instead of load_pil_baw() since...
• f211ab66 - [data.montgomery] Update documentation reflect more truthful information about dataset

Compare with previous version

marked this merge request as draft

@mdelitroz: something else that came into mind - if the Shenzhen dataset is already RGB (even if, in reality, it is not), then we should load that as is and not convert it to grayscale at raw-loading time. We will save the RGB conversion for models that need it.

added 1 commit

• 249274a9 - We decided to handle the resizing of the input at the model level and not at...

Compare with previous version

added 5 commits

• a4686009 - [models.pasa] Set grayscale/resize model transforms
• 8361e117 - [data.montgomery/shenzhen.loader] Simplify loaders, adjust some variable...
• 3b5fba10 - [data.datamodule] Double-check model_transforms are set before datasets are...
• b8f0c8cd - [data.montgomery/shenzhen] Simplify documentation maintenance
• 7e752335 - [data.datamodule] Fixes cached dataset model transform application; Improves...

Compare with previous version

I added some commits to simplify certain aspects, document more, log more and to fix a bug in the CachedDataset, where it was not applying the model_transforms...

marked this merge request as ready

mentioned in commit d0b7d8b8

merged