How can run `nlg-eval` to also produce ROUGE 1 and ROUGE 2 scores ?
How can run
nlg-evalto also produce ROUGE 1 and ROUGE 2 scores ?