Hi I would like to propose a better implementation for 'test_indices':

We can remove the unneeded np.array casting:

Cleaner/New:
test_indices = list(set(range(len(texts))) - set(train_indices))

Old:
test_indices = np.array(list(set(range(len(texts))) - set(train_indices)))

Having problem with polish letters. universal-data-tool (web version and windows app) is splitting words if there is a polish sing (utf-8). And treath is as an distinct word. In example from .png photo "potrzebuję" and "różyczkę" are single words.

    "\U0104",        #Ą
    "