Hi Patricia,
we have developed a web interface for creating ground truth, i.e. transcribing pages line-by-line which you can find on Github [1]. Since this is an important part in the process of training and evaluation and can also be used for other OCR engines that produce hOCR like tesseract, we figured it made sense to spend some time in an ergonomic user interface.
But for training, text detection etc., you will need to use the CLI tools or the docker container [2].
Best
Konstantin