Which file should I run for each script?

[shakiba-bakhtiari]

I read the ReadMe file and followed the requirement steps. However I have some CSV files that each one has some columns that differ from other files. Which one I use for executing parts of this code?
(for example the combined-newsqa-data-v1.csv file has "story_id", "story_text", "question", "answer_char_ranges", "is_answer_absent", "is_question_bad" and "validated_answers" columns and in split-data folder the train.csv, test.csv and dev.csv files have "story_id", "story_text", "question" and "answer_token_ranges " columns.)
I would be forever grateful for your kind help.