Working on preprocess and predict container

Finished manual tutorial

Removed debug prints

Working on setting up automatic weight downloading

Small changes to BIDS scripts

Finished first version of scripts necessary to finetune models

Finished functionality for predicting via python

Working on docker containers

Cleaned repo for new setup

move the argumentparser and execution into a function main()

Move the script to predict into a new module cli within the package

Tried moving data folder (although I don't want to)

Minor modification to data import

Tried to add data folder to packaging

Updated docker containers, small linting fixes, fixed bug with weight repository and directory

Added notebook for finetuning brain age model

Finished READMEs for now

More tests and comments, modified README

More documentation

Added unit tests for the weight repository

Added a notebook for running SFCN for binary classification

Created first docker container

Finished models, pre/postprocessing and example from paper 1

Clean start for new version

Updated unit tests

Added contrast augmentation

Bugfixes in augmentations

Minor bugfixes

Added crop augmentation

Added more augmentations

Added multisample dataset for datasets with multiple session per subject

Bugfixes

Added noise to augmenter

More performance tests for augmentations

Added multiple augmentations and started on performance tests

Started working on additional augmentation functionality

Added code for nested crossvalidation

Finished top level script for crossvalidation and implemented tf dataset in fit_model

Refactored and added some new top-level scripts

Finished VGG11, added sliced metrics

Finished VGG11 code, added script for learning rate sweep

Fixed restrict_range for regression models

Merge pull request #4 from estenhl/transformers

Finished 3D ViT

Added onehot encoding to categorical labels

Finished demo-code for Vision Transformer

Added functionality which is necessary for training domain-adaptive models

Minor changes to domain-adaptive models

Finished infrastructure for training domain-adaptive SFCNs

Changed setup for encoding variables in multilabel dataset, changed naming in binary labels