Predict multiple sentences for Textual Entailment

How can I predict multiple sentences for Textual Entailment, so that I can save response time?