training for a new font from images

[Ghinwa Choueiter]

Hi there,

I have collected thousands of examples for my font. Each example is one line of text. the images are png RGB. I ended up not doing any preprocessing whatsoever to the images during training. Is that ok? 

If it is recommended that I do pre-process, then what are the suggestions?

thank you,

G