I have trained a LDA topic model on a corpus of 20,000 documents (training sample). I want to use the trained model on a new sample of 2,000 unseen documents and generate to
