Justus Schock

Add Multi GPU support for TF Graph and TF Eager backends [Distributed Training in TF](https://www.tensorflow.org/beta/guide/distribute_strategy#using_tfdistributestrategy_with_custom_training_loops)

Metrics

1

Scikit-Learn metrics often don't support enough dimension (i.e. they often don't support 3 dimensional data (NxHxW) or even 4 dimensional data (NxCxHxW) and thus we either need to provide custom...

We should add additional tests for the following things: - [ ] `Resume` function of the `Experiment`

Recheck dependencies for deprecations

Fix coverage calculation in CI/CD

Docs

1

Check automated doc building (on github page via CI/CD and on https://delira.rtfd.io

[FeatureRequest] Dataset Metrics

5

Find a way to provide a lazy way for calculating dataset metrics since they'd currently break the intention of lazy datasets (see #66 )

Integrate hyperparameter tuning with ray

Unload image

4

**🚀 Feature** Possibility to "unload" an image to save memory **Motivation** When working with huge datasets, it can easily happen that one runs out of memory. To avoid this, it...

Repr should not load the image

5

**🚀 Feature** Changing the way `__repr__` gives an representation of the image. **Motivation** Debuggers call `__repr__` to give a preview of the python object. **Pitch** When having torchio Images that...