We have two csv files, one containing hindi words, another containing marathi words.
The data format is:
id, word, gloss, example
Words of two different languages are cognates if they have the same word id.
We create non-cognates from these two data files by running mismatch.py