I use the below code to create the tif, gt, txt, and .box files. so I need to know what the best scenario of xsize/ysize of. how many words should be in per line on the training_text file? if I make a long line and set xsize/ysize as I need to cover the text can affect any issue? or tell me the ideal size of it? thx in advance.
subprocess.run([
'text2image',
f'--font={font}',
f'--text={line_gt_text}',
f'--outputbase={output_directory}/{file_base_name}',
'--max_pages=1',
'--strip_unrenderable_words',
'--leading=36',
'--xsize=3500',
'--ysize=500',
'--char_spacing=1.0',
'--exposure=0',
'--unicharset_file=langdata/ben.unicharset',
])