2 questions about training tesseract

[LHW]

(My english sucks, so please understand me)

Hi, I'm studying how to train tesseract with tesstrain

I have handwriting/printed font dataset(.tif and .gt.txt pairs).

I read tutorial, and I successfully made new traineddata file, but I want to add my dataset in existing .traineddata file. (ex. eng.traineddata + dataset)

1. can i add data in traineddata file that already exist? or can i merge two or many traineddata files in one file? (except [-l model1+model2])

2. There are syllables, words and sentences in my dataset. can i put them in one folder and train? or should i train each of them seperately?