You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
My first file had very bad OCR in the pdf (english letters for the lao language). It seems to influence the ordering model in some way? If I use surya with the same file but no OCR in it (print to file before), then the result is correct:
I've had results that seem just reversed, numbering from the bottom of the page to the top while it should be the opposite.
I realise I should have posted on
maker
, I use:The text was updated successfully, but these errors were encountered: