Creating training data for a language with a complex name, like ita_old or chi_sim_vert

26 views
Skip to first unread message

Sim Tov

unread,
Aug 18, 2021, 6:40:05 AM8/18/21
to tesser...@googlegroups.com
Hello,

I try to create training data for a language with a complex name similar to ita_old or chi_sim_vert. However when I run the command:

tesstrain.sh --lang eng_old  --fonts_dir ....

I get this error:

=== Starting training for language 'eng_old'
ERROR: Error: eng_old is not a valid language code

How can I cause tesstrain.sh to accept 'eng_old' the way 'ita_old' is accepted?

Thank you in advance!

Quan Nguyen

unread,
Aug 20, 2021, 8:27:09 PM8/20/21
to tesseract-ocr
Pick a name that it accepts and then rename the output file to desirable names.
Reply all
Reply to author
Forward
0 new messages