Evaluate machine translation output against reference: WER, PER, TER, BLEU - View it on GitHub