Glove on multipe documents

[nayangarg]

I have a corpus of multiple documents.
How to fit the model on such a corpus.

I read somewhere that if I create a single document with all documents separated by an empty line, it would work. What to do when we have documents where there may be empty lines in between?