Unlocking Anticipatory Text Generation: A Constrained Approach for Faithful Decoding with Large Language Models

Open in new window