Otbench

Release version `2024.1.7` of `otbench`
* migrated from scikit-learn `r2_score` to a coefficient of determination computed using `scipy`
* small edits to overview notebooks
* improved contribution guidelines

Future improvements:
* add a new `usna_cn2_full` dataset with all available data across all sources, and without interpolation

Release version `2023.11.28` of `otbench`
* improved deep learning benchmark modules
* added new notebooks to provide an overview of implementing dep learning models for optical turbulence strength prediction and forecasting.
* updated some model names
* re-ran benchmarks, persisted as `expiraments.json`, with the new `RNNModel` included

Future improvements:
* add a new `usna_cn2_full` dataset with all available data across all sources, and without interpolation
* improve repository with discussions and templates for contributing and issues

Release version `2023.11.14` of `otbench`
* added support for deep learning models through `PyTorch`
* improved test coverage
* updates overview notebooks for regression and forecasting

Future improvements:
* make all baseline models deterministic in their scores
* add a new `usna_cn2_full` dataset with all available data across all sources, and without interpolation
* improve repository with discussions and templates for contributing and issues

Release version `2023.10.27` of `otbench`
* first public release
* improved test coverage
* full README coverage
* deemed production ready for atmospheric research and benchmarking

Future improvements:
* make all baseline models deterministic in their scores
* add a new `usna_cn2_full` dataset with all available data across all sources, and without interpolation
* improve repository with discussions and templates for contributing and issues