These are excellent -- are there recommended settings to go along with the trained data?

[virtuald]

The scripts in this repo have been incredibly useful, and I've found it really easy to adapt to my own training text/font choice. Thanks so much!

When using a known (ascii) monospaced font in a known orientation, are you aware of any settings that I can put in a config file that will help tesseract get better quality. In particular, the wiki mentions that disabling the system wordlists helps if you're trying to not recognize sentences -- but that doesn't really seem to help at all.