Enhancing Abstractiveness of Summarization Models through Calibrated Distillation

Open in new window