Fine Tuning for ± a few characters in jpn and sim_chi

[Hoang Vu]

As the wiki say :

New feature It is possible to add a few new characters to the character set and train for them by fine tuning, without a large amount of training data.

I'm trying to add some symbol to jpn.traineddata  by using fine tuning a few characters but i'm so wondering how many  training text line is good for japanese?

In english eng.training_text file have only 70 lines of text but in japanese this have 1670 lines of text. And i think if i add all of 1670 lines to training text i maybe got overfiting.

Must i have put all of this to  traning text or a few of lines?

https://github.com/tesseract-ocr/langdata/blob/master/eng/eng.training_text

https://github.com/tesseract-ocr/langdata/blob/master/jpn/jpn.training_text

Thanks for read!

Sorry for my bad english !