Release 0.1.7

- fixed wrong MD5 checksum in a dataset
- removed dependency to LibMR
- fixed citations
- added checks to segmentation metric calculation

Release 0.1.6

- Fixed Bug in Fooling Images
- Fixed Bug in Temperature Scaling
- Fixed Devide Mismatch in KNN Detector
- Improved MCD Implementation and Documentation
- Added NINCO Dataset
- Updated README

Release 0.1.5

* Fixed but in torchtext examples
* Added contribution instructions
* Added DICE Detector
* Added ReAct Detector
* Added SHE Detector
* Added Temperature Scaling Detector
* Added RMD Detector
* Added ASH Detector
* General Documentation Improvements

Release 0.1.4

- updated torchmetrics
- improved MCHAD and CAC examples
- improved README
- improved setup.cfg
- added datasets and benchmarks for OpenOOD

Release 0.1.3

- added ODIN benchmark
- migrated to github actions

Release 0.1.2

- fixed bug in CAC Loss
- added example for MCHAD Loss
- improved Mahalanobis implementation
- improved documentation
- improved Deep SVDD implementation
- fixed bug in FoolingImages dataset

Release 0.1.1:

- documentation improvements
- added feature based interface
- bugfix in fpr95
- removed deprecated ViT

Release 0.1.0

- Implemented Entropy Detector for Classification and Segmentation
- Implemented entropic loss for segmentation
- Added examples for OOD detection on text
- Added example for supervised OOD segmentation
- Updated Documentation and Readme
- Minor code refactorings
- Fixed device handling in Energy Regularization Loss
- Implemented metrics for segmentation

Release 0.0.16

- documentation improvements
- example for segmentation model added
- fixed bug in energy regularization
- fixed bug in objectosphere

Release 0.0.14

- General documentation consistency improvements and fixes
- Fixed inconsistency in Chars74k dataset
- Fixed bug in CIFAR10C dataset
- Implemented PixMix
- Fixed bug in KL-Matching
- Improved integration with pytorch-lightning