Finetuning with images

[Mox Betex]

Can someone help me how to do finetuning with images in OCR-D.

I have tiff/txt files in ground truth, but not sure what to do next.

I want to fine tune srp language (Serbian cyrillic).

[Mox Betex]

I have specified srp as START_MODEL in makefile:

# Name of the model to continue from. Default: '$(START_MODEL)'
START_MODEL = srp

Is that all that I have to do?

Training did start without problems, but I am not sure is that all that needs to be done.