* XLM-R now passes common tests + Integration tests * Correct mask index * Model input names * Style * Remove text preprocessing * Unneccessary import
* add first files * add xlm roberta integration tests * make style * flake 8 issues solved